Hi Anton, I'm a chemical biologist. There is an entire class of proteins called chaperone proteins which includes heat shock proteins (e.g. HSP60). These proteins assist unfolded proteins to fold the right way to ensure proper tertiary structure. These evolved as proteins got bigger and more complex. I work on an amyloidogenic disease which has dysfunctional chaperone activity as a component. Obviously the first spontaneously formed proteins/enzymes didn't have the luxury of chaperone proteins, but they were likely much smaller and their function much simpler.
It may be astronomically unlikely to have a protein with the same exact sequence that folds in the same way, but many amino acids within proteins (depending on the protein) can be changed without affecting the protein's function. This makes the possible number of configurations that can have near identical functionality while having different sequences rather significant.
The fact that evolution doesn't have an end goal, what we're made of doesn't necessarily mean other life has to be formed similarly, or that proteins and DNA are the most efficient basis for life. We can only base our concept of life on what we see, not what we're likely to find.
And there are also many, many more proteins that have similar enzymatic functionality that have completely different sequences and structures. This really seems like an argument from ignorance.
The man is a superhuman lol, putting out videos everyday while working on a PHD plus all the day to day stuff everyone goes through, and dealing with a tragic loss, he's definitely something special and our world is better for it.
I have such immense respect for him. I can tell he's holding onto so much emotion. He's my favorite youtuber because of his strength and dedication to his passions.
Students aspiring to become “Doctors” pursue an MD; medical researchers / professors go the MS or PhD route. I only clarified the distinction because a lot of academics / doctors get rather (annoyingly) pedantic about their (MANY) abbreviations.
The solution to this paradox is that during the folding process, proteins pass through a series of INTERMEDIATE states that reduce the number of POSSIBLE conformations. These intermediates are like ANCHOR points that help direct the protein to its "CORRECT" structure. In addition, hydrophobic interactions also play an important role in the folding process.
@@BernardoSaab Deterministic. Look at those drawings of protein structure. See the long coils? Those are extremely energetically favorable. That alone eliminates most of the other 'possible' configurations.
Agree, additionally conditions (e.g. temp, pH, salinity, redox potential) also influencing the folding process further limiting the possibilities and as such speeding process. In modern organisms there are also other proteins, chaperones, some of which even further limit the possibilities and speed up the process. The growing protein chain is "anchored" in ribosome limiting one end movement, and reducing options. In prebiotic Earth similar role could be played by some mineral material, like particular clay. The refolding (denaturation and folding again) works also in vitro, crystallographers use it many times to obtain protein for their studies - the process is based on e.g. slow redox potential change in particular physico-chemical conditions, it usually takes about 24-48 h and vast majority of the protein is precipitated because of incorrect folding, yet still many molecules gets folded. All the factors mentioned here and by You others folks has tremendous influence on the limiting the "options" that amino acid chain has and speeding the folding process. Calculating influence of all of these factors would be a nightmare, bigger that forecasting weather though.
I think of it like origami - you have to fold creases in your paper at various points in the construction of the model, otherwise the paper won't cooperate with you and it's very difficult to get it to fold correctly. But once you've made the creases, the paper folds into place easily.
Yes. I think this is the same kind of argument that can be used to refute creationists who claim that evolutionary theory cannot possibly be correct because how could a structure so complex as the eye be formed by a series of random mutations all happening in the just the right order. The response is the order of mutations is not random, but rather the first mutation limits the number of possible second mutations, and then the second mutation limits the possible third mutation and so on.
This is a rare case where I disagree with you, Anton. For a start, the problem is predicting the shape given the sequence of proteins, not the unordered bag of them. (That still made it one of the hardest scientific problems we know.) But the proteins don't then fold randomly, physical forces (like electrostatic attraction) act all along the chain, causing folds to start at various points. As they fold, new parts of the chain come close enough to affect one another to start fresh folds of the already-folded sections. So there's a logarithmic process at work, as well. Because AlphaFold 2 predicts the structures with nearly the same accuracy as experimental determination, and because proteins do tend to fold predictably in nature, it means the process cannot be a random one, it must be following physical laws that aren't overly sensitive to initial conditions. So in that sense it's similar to predicting the path of a ball rolling down a surface under the force of gravity (just with the parameters of the problem scaled up hugely).
I completely agree. I would also say the claim that most of the folds "do nothing" is a bit misleading. A differently-folded protein may have no function within the organism that produced it, which is what we might expect. The various chemical components within the organism coevolved, so to speak, and can't be expected to interact meaningfully with just any random particle (outside of something like a part of an immune system that is built to capture/destroy anything it _doesn't_ recognize). However, a "randomly" structured protein can still have chemical interactions with other specific particles, leading to its own complex systems of self-replication and adaptation. If randomness does play a significant role in how life initially develops, it could be that the biology on this planet is extremely particular. In that case, if we find aliens we would expect their biology to use all different proteins, or perhaps a different mechanism entirely. Protein folding is really interesting and there is a lot we don't know, but I think claiming that it may be the primary solution to the fermi paradox is somewhat analogous to claiming that non-Chevy vehicles are very unlikely to exist, since a Chevy won't work properly if its ECM (computer) is programmed even slightly differently.
Very well stated. I'd like to add, as an aside, that there is (probably) no such thing as 'randomness' in the universe, only hidden variables, so small and numerous as to give the appearance of 'randomness' for their incalculability.
The issue I had was with the 10^300 number from Levinthal's Paradox in the first place. Even if protein folding was randomly determined it's very obvious the probabilities of a specific shape change depending on the environment and the paradox doesn't take that into account at all. Of course you are going to get a excessively large number if you assume proteins being created was a random incident, but that is a current debate in origin of life research and there are many who think evolution began before the first nucleotides even. If you assumed all of the cells in a human body assembled randomly then you would get a similarly ridiculous number, but we didn't assemble randomly we evolved over the course of a very long time. It is feasible that the same would be true of proteins.
I also want to add --- protein fitness landscapes, judging by what we've explored of them so far, rather strongly suggest that if you took any protein in nature and folded it an entirely different way that it'd still have *some* chemical function that could possible serve a living thing. I think this Paradox is built on fundamentally flawed math in numerous ways, not only be not accounting for the fact that some ways of folding have more likelihood than others, but also, that proteins need not be that specific to perform *a* function. This *might* be an issue with certain basal, universal functions one would need, but these functions would not even be needed in protolife, necessarily. (Judging by the likelihood of RNA theory, probably all you need is reproduction of a single macromolecule to kick start life, and the recent experiment that produced autocatalytic RNA suggests that the search space of RNA structures that reproduce themselves is possibly large enough compared to all possible RNA structures that a planet that produces RNA in bulk will inevitably produce one --- I mean, we produced one by producing RNA in much, much smaller bulk... in a single lab.) I would suggest that at best, this merely suggests that all the possibilities there are in what functions life can achieve is even more open ended... not like it wasn't insanely open ended in the first place.
@@Rampart.X Have you followed the experiments over the years that tried to invalidate the proposed lack of hidden variables in quantum mechanics. They have all failed. There's a lot of evidence for randomness in quantum mechanics, and none for the hidden variable hypothesis any more, as I understand the current state of things. The only thing still supporting hidden variables is human uneasiness at an intrinsically random Universe.
Anton so many of us are all glad you didn't become a doctor. Me personally you have taught so many things that have lead into learning so many other things in several fields. I have been able to make connections in society and my surroundings that I never would have if I hadn't watched some of you and other RUclips Teachers videos. Seriously man Appreciate you!
Appreciate you Anton. Your videos are always a delight. Sometimes it's just nice to hear the "hello wonderful person" and be reminded it's our ability to learn and wonder at the mysteries that often makes us so wonderful.
Hey dude, really sorry to hear about your mom. Losing a parent is always really hard. Though I am happy for you that you were able to use that tragedy to reevaluate what you wanted out of life and chose to pursue what makes YOU happy. Proud of you buddy
Anton, you have become the premier science explainer on YT, perhaps all of the internet. I am constantly impressed by your diligent attention to detail. You represent the very best of the world of digital communication.
If it's predicted that it should take the age of the universe to fold protein yet it only took a few milliseconds then that in itself could mean something intervened or sped up the process or we are just missing something vital to understanding this
I remember when I was still a Biochem student my professor strictly warned me to not try working on optimizing rubisco or any rubisco research. He said that there have been many who have wasted their whole careers in doing so, and that if evolution could have optimized rubisco it would've found a way to do it by now.
@@baronka3404 I'd argue that a lot of things in nature _are_ optimized. At least up to some factor of "perfect". For one, genetic search algorithm in computing basically mimic natural selection + mutations to find optimal solutions to a problem. And they are pretty good at it. Nature basically had thousands and thousands of years to optimize stuff. Things that were better lived and reproduced more, things that didn't work died off and got culled. Secondly, and probably most importantly, things in nature are probably optimized using criterion that are unknown/different from what we think of. Optimizing one thing almost always come with a cost somewhere else.
Your mum would be so proud of you Anton - you are sharing info with the masses that is so important - we are glad you decided to change fields! Thanks for the video - I learnt a lot! Love from the UK.
At 6:40 Anton, when you are discussing whether Ribisco could evolve as the same protein molecule on another planet and function the same way elsewhere, it puts me in mind of the two main types of protein molecules that can carry oxygen in blood, and in this sense it is two different elements that are employed to do that job. In most animals it is Iron in Hemoglobin, but in Octupi it is copper. So even here there is at least two known ways on Earth, of carrying oxygen to and waste gases from cells via the vascular system of of animals.
This remark appears to be in the same lines as my initial (layman's) thoughts. The combined facts that there are 1E+300 ways to fold, and that if we change one thing in what we know is working, it won't work any longer, does definitely not validate the conclusion that all of the other (1e+300)-1 ways of folding will do nothing to support life. Also, 1 in 1E+300 seems way, way to large to describe the odds of any conditions that are needed for the functional fold. But I have to admit that this is based on intuition.
Channichthyidae, sometimes called crocodile icefish have translucent blood, white hearts, and have somehow adapted to live without red blood cells or hemoglobin. Their blood is cloudy with other proteins and macromolecules, some of which help to keep the blood functional in the cold polar waters.
@@Nerthos "which sadly leads to their poaching for pregnancy tests" Also their blood can be used to detect other things, so the poor things get "farmed" for it.
Proteins don't fold all at once. They are emitted one amino acid at a time from the ribosome. So the first part will get to fold long before the last part gets assembled. This means there are many fewer possibilities than in the all-at-once model. Because it's easy to add more DNA at the end, it also means that the core functionality (of say an enzyme) will tend to be emitted earlier, with the later parts of the protein wrapping around it and providing support structure.
You beat me to it. The "solution space" is thereby very much constrained. Also, apart from being produced sequentially and folding sequentially along the way, kind of like the silicone worm extruded out of the tip of your cartridge nozzle, or the shavings coming from the tip of your lathe's cutting tool, there is possibly "instrumentation" present in the intracellular environment, helping the protein folds favour particular macro-shapes. Someone's just received a thousand upvotes for mentioning the Chaperone proteins... voila, some of it is known and has a name :-) Wondering "why does it fold exactly this way, if it's just a pile of amino-acids, with a 10^-30 probability" is a bit like asking "how does a complex piece of software happen, say the MS Windows, consisting of some 30 letters of the ASCII character set". Except that here, unless you prefer to believe in some $DEITY, there is no genius creator involved, writing the code. The cellular protein factory and the DNA have co-evolved for a long time, driven by the ultimate in evolutionary genetic algorithms. Over billions of years, and possibly billions of generations, fitness for survival has been the ultimate "scoring function" = the judging force over what stays in and what goes out, of the genetic sequences that end up in the folding proteins being produced. This is how the endless space gets searched for the viable paths.
Yes. Thank you. I only learnt this recently and it makes AlphaFold far less impressive. Impressive it still is though. I think DeepMind would prefer you think there are really 1e+300 possibilities
Your right, however when a protein is melted, via heat blasting it's quaternary, tertiary, and even secondary structure apart, very often they do properly reform when they are cooled down. Certainly not always, but often.
adding to that, there are also so called chaperones, proteines that assist in the folding process of many proteins. Not to mention that amino acid chains are not like a rope which you can fold any way you want. The side chains of the various amino acids have different properties. Some will attract each other, while others repel each other. Then there are restrictions to how much you can twist a chemical bond. At least he mentioned things like ph and temperature affecting how proteins fold. It's all about the forces that act upon each other. Those forced attract or repel leading to the final structure. So there are many restrictions in place, which greatly reduces the number of possible ways a protein can fold. Proteins can fold in many ways, but in the given conditions and the way they are build, there is only that one end result. And the existence of the chaperones is just a result of evolution. There is nothing paradoxical about it, what works works. If it works it can survive and pass on. If not, then that's that, it will not survive or at the very least no procreate. Life doesn't have to be perfect. It isn't. It just has to work somehow.
Love your channel Anton. I never realized that you lost your mom eventually decided not to become a doctor but do remember you had background in biochemistry. I had sort of the reverse situation in a way. My dad died of ALS when I was younger, so I had decided to become a doctor, but eventually I realized this was not a pure motivation of mine, but was fueled by revenge and my inability to accept death. Eventually I decided to become a math major when I entered college, and rediscovered that childlike wonder in it, unclouded by grief.
Yes....pursue whatever created that child like wonder, that giddiness....whatever that field is for you, you can not go wrong ! Its anatomy/physiology for me. I feel like I'm entering a magical cosmos and I need to understand all of it, and it makes me happy. (Then I can't shut up about it to everyone I'm close with. 😅) Stem fields unite!
I am glad you chose another path, not just because I get videos out of it, but because I have seen many people following a path set by their parents at their own detriment.
If I remember correctly one of the leading hypotheses out there for how proteins fold so quickly is the energy landscape funnel model where the protein initially samples a wide range of conformations but as soon as it becomes slightly more folded the conformations it can sample become drastically limited, kind of like cutting off entire branches of a decision tree. This is in line with our understanding of thermodynamics with a highly flexible state being an unstable one for most proteins (now intrinsically disordered proteins are a whole other story) and the experimental evidence of protein (un)folding in domains and through specific stages. The funnel hypothesis also explains why some proteins are prone to aggregation and how refolding a protein is possible but needs an external boost like a chaperone protein. Great video though, love your content!
Exacrly... proteins do not need to try all positions. They minimise their energy. The same as a marble that rolls down a hill... there are infinite many ways, but the marble will follow just one (at least if it is a macroscopic marble). The trick of living nature is, to control this process by shaping the landscape (through the environment like pH, ion concentration, fatty membranes in the vicinity and the process itself with varying synthesis soeed and chaperone helpers around), so that the marble roles down a useful path and arrives at a designated target point and not elsewhere. When a protein gets unfolded and refolds on its own, then the way down of the marble is more random (regarding the final position) - or maybe not random, but not useful anymore. That's when protein turn nonfunctional. It depends on the sequence. So Levinthal's paradox is none... no superpower or superintelligence required, just the laws of physics and chemistry. And a lot of trying and improvement over time to get there where nature is right now. Cheers
pushed by the surrounding environment, hydrophobic collapse (hydrophobic parts of the chain flip inwards and are shielded from the water solvent by more hydrophillic parts) is one of the proposed mechanisms for eliminating a lot of "wiggle room" very quickly, if I remember correctly. I could also imagine that there is some form of folding assist in cells where enzymes catalyse conformation changes but also induce conformation in the first place.
This makes a lot of sense to me, depending on conditions around a protein chain, all the astronomically large number of final configurations it can reach are possible but they are not all equally likely. It's likely a continuous process where at every new step along the way, there is a sampling function evaluating the current conditions and a folding function directing the next step in configuration based on a more limited set of possibilities. It also sounds pretty close to how I imagine alphafold comes up with these predictions although I don't think that accounts for a continuous environmental evaluation at every (infinitesimal) step along the folding process.
Although machine learning models like AlphaFold aren't mechanistic, they can still be useful for understanding how folding works. For example, because the input for AlphaFold is the amino acid sequence, its success at predicting folding structure implies that the sequence encodes the relevant information needed for folding. Advances in neural network interpretability may also provide insight into what data correlations the algorithms are leveraging to make predictions. Very cool!
I think it rather implies the opposite - that there is extrinsic information (which we don't know yet) not encoded in the sequence but causing the recognizable patterns. If you thoroughly analyzed the entire U.S. census data minus 1 name and then were handed a pile of letters, "nhismjh to", you could probably predict with good odds that "john smith" is a likely name. At least more likely than "otihshmnj ". But that's not encoded in the letters themselves. It's extrinsic things in our language and history and how we form names. What's really cool is that, for things like this (protein folding) we now are developing AI that can do the analysis, recognize the patterns, and predict the outcomes. It can't yet tell us what the extrinsic causes are or why, but it's still a huge step! Could dramatically reduce the area we need to search to find the answers. Exciting stuff!
@@rascta Yeah, that makes sense. I also don't think it's a matter of the sequence encoding anything as far as a meta instruction or something; I think it's a matter of only these specific arrangements actually occur in our conditions here. So the ML is essentially ruling out all of these "possible" arrangements that never actually are seen in nature (at least in our conditions) simply by having never observed them and thus not having them as options to arrange as it predicts the outcome. Those conditions include the ph, temp, etc like Anton mentioned, but also the other proteins that are interacting with it during folding, and other catalysts and potential limiting factors we aren't aware of yet. So we have a lot of fundamental physics happening, which leads to chemical paths of least resistance as bonds try to form in the most energetically favorable arrangement, but that is counter balanced by the direct forces imposed by the local environment (i.e. chaperone proteins), and this leads only a tiny subset of useful proteins ever sticking around and not being reabsorbed by other processes.
Would imagine that the AI is doing pattern recognition. It is recognizing combinations and sequences and eventually we will make another AI to tell us what the other AI did.
Folding is SEQUENTIAL. Proteins are produced ONE amino acid monomer at a time, hanging from the assembly point of the ribosome. The first monomer has nothing to fold against until the second monomer is extruded. Now there are two monomers flopping around, but one of them is still stuck to the ribosome. Of all of the possible ways they can flop together, one of these ways has the maximum stability (lowest energy), so the monomers stick together in this most stable configuration and tend to STAY that way. Next, monomer THREE is extruded and comes into contact with the grouping of the previous TWO in the prior maximally stable configuration. So the new monomer is bumped into by the TWO previous monomers, already folded together, which bounce around and form the maximally stable configuration CONSISTENT with the previous folding. There are NO FREE CHOICES, because, at each stage, only one configuration is maximally stable GIVEN the configuration in which the prior monomers have folded. Hence, at each stage, there is only ONE configuration possible. Hence, THERMODYNAMICS drives the folding process. It does not matter if other configurations are possible, even if some of them are more stable than the current configuration, because the previous configuration is already locked into place. I am ignoring the complexities introduced by chaperone proteins, which are already out there.
One slight extension of this theory: all possible minimas are expected to occur, just at different probabilities, there is not one choice only, but many, with one being the preferred one in most cases. This still explodes in combinations, but vastly reduces the space and especially makes for a much more compact space of probabilities.
Thank you. And such rules apply throughout the universe wherever a cosy, long-lived and relatively stable planet orbits. Brings down the odds against rarity. The eukaryotic jump does seem a rarity, but that’s another problem. Should be more variants.
1. Why are you using all caps? 2. Does it even matter? Why would we think any sequence we identify has to be some kind of unique formula for success toward intelligent life? No biology degree necessary to realize we can’t begin to conclude there’s only one way for it to happen.
Actually, when it comes to polypeptides over a certain length, the kinetics of the cellular environment interfere with proper folding, thus the necessity for isolation inside the chaperone capsule. The nascent protein is literally transported into the chaperone where it achieves its tertiary structure. Even so, the kinetics of such a long chain allow for multiple bonding configurations separated by miniscule energy levels. The Levinthal paradox still remains poorly understood.
Thank you Anton for keeping my brain active and engaged at my age. Although I could never do or understand the science/math behind your videos, your explanations are enough for me to get the concepts you are trying to convey. You bring a sense of wonder and awe to my existence through science.
Seems that way too for a knucklehead that spent too much time with the kids in math for the gifted and talented getting them to work together to write code that could have been created by this that could not, alone. So fun. Good times.
I think it shows good character that you chose to take your own path out of the tragedy. You are educating thousands of people including me, so thank you for what you do. Your ability to make greatness out of life in the face of loss is inspiring to me.
Hi Anton, I am a 58-year-old chasing his Ph.D. At present, I have two master's degrees and am waiting for word from the programs I chose. I wanted to encourage you. Chasing a Ph.D. you want may turn out to be the exact opposite of your previous Ph.D. experience. You are very talented. I have often thought of you in my own removed storyline about you as an astrophysicist or an astrobiologist. I hope you can turn all your efforts into an amazingly self-fulling path! Best wishes!!!!! Rob
We do have a pretty good idea of how most of this is done though : chaperone proteins are shaping proteins as they are synthesized. So while proteins will by themselves fold into a stable configuration, they are also shaped by the proteins they are synthesized around (which is determined by a sequence in the mRNA, which will lead the mRNA to be transported in a particular place in the cell, which may vary with epigenetics). This makes things a bit more complex than just knowing the sequence of the protein to determine the shape, the context matters. Knowing the sequence of the mRNA and its epigenetics, on top of the whole machinery of the cell, is what is really necessary to determine the shape of the protein.
The video isn't talking about the folding of proteins in a cell, it's the formation of said proteins in nature before any life exists. Chaperone proteins don't answer the question they're a part of the question since they must ALSO form spontaneously at some point for any other proteins to use them as a guide. How did the exact proteins needed to CREATE life form on their own randomly and why not anywhere else......not how does life on earth create the proteins it's been creating for eons. Two different questions.
@@MrBottlecapBill I guess we can only assume, random effects eventually create something complex that carries on being more complex. Pretty strange effect.
@@MrBottlecapBill That's my point : while the very first proteins did form spontaneously and folded uniquely based on electrochemical properties, everything that formed after formed in an environment with preexisting proteins interacting with the formation. Any protein that is common to all life on earth comes from dna being translated, not from spontaneous formation. So the mechanisms have had the time to evolve quite a bit and are nothing like the primordial soup. The environment matters a lot. (as for why nowhere else : we have no proof life never emerged on our closest neighbours in the first place, so let's not jump to conclusions : for a while there was no life on Earth either, and at some point it will disappear. Time and the lack of data may be the main reasons why we didn't find extraterrestrial life yet, not even talking about a civilization like ours which is going to burn all of the resources of the planet in a few centuries. That would be looking for a needle in a haystack while not being able to feel your extremities and being blindfolded, that's about how bad we are)
@@MrBottlecapBill Assuming a random event on the nature where the protein doesn't even can exist doesn't sound very intelligent... the video states a random chance, who said the folding is truly random? have they acknowledge EVERY interaction? atoms and molecules have physics and chemical laws to obey, it's never random if you scrutinize enough.
I think this paradox can be slightly resolved by recognizing that there may be many potential pathways that life can use for a given function. So initial production of a specific protein might be highly improbable, but reproducing a protein that achieves a similar function could be much more likely.
==== Even getting two proteins to work together is beyond explaining. And ten is worse still. No way the folding protein process originated from dirt. Please explain how this could be? ==== Evolution = Self Assembling Atoms = Impossible. ===
Imagine that each environment where life is POSSIBLE (which we still don't know all of them, because we keep finding life in more and more extreme environments) has it's own set of one, two, (three, thirty, six thousand, or more?) "shapes" which a protein CAN fold into which would support life in that specific environment. To make it easier (than trying to think about all the huge numbers of different ways proteins can fold) imagine that these shapes are a 2D silhouette for combinations of Legos which is cut out from a board. You have a box of random Legos provided by the environment (the environment favoring some Lego blocks more than others, as per it's unique environment). Your task (and the task of about a bazillion others at the same time) is then to find just ONE combination of constructed Legos which can EXACTLY pass thru the cut-out shape. Of course, there are many different combinations which will likely work... but you only NEED to find ONE. Now let's go back to that huge number of 10^300... That's how many different ways the entire set of Legos CAN be put together. But you don't need to go thru each of them one at a time... because biology and chemistry almost never do things like that in series; it works in parallel, so it can try many, many 'shapes' at the same time. In fact, it also favors things that are easier to try and going toward more complex as the assembly continues on. So you have an environment which favors a subset of proteins, which already cuts down on the number of variations... And you have more than a single attempt at any given time (assuming these amino acids are in an environment where protein assembly is possible - aka: "where life is POSSIBLE" )... Instead you have bazillions of iterations being tried at the same time. Suddenly this "paradox" is looking like much less of a paradox, yes?
Something I wanna add is that not all 10^300 permutations are equally probable. For example, hydrophobic amino acids are more likely to interact with each other than hydrophilic ones (and vice versa). This discards a lot of permutations that simply aren't as stable or possible to form in the first place. Therefore, intermolecular forces such as Van der Waals or hydrogen bonds may induce/promote specific permutations. So, is possible that a lot of unstable states collapse into more stable ones. In general, the system will keep collapsing to more stable states which leads to the desired folding state.
@spacekoala2887 Water certainly must have a major effect on protein folding. But the cytoplasma in a cell contains somuch more, which also should affect it. I guess that chaperon proreins doesn't even has to be for only that purpose.
Directing the infinite permutations of protein folding into adaptive structures might be the actual intelligence behind intelligent life - not iPhones and Dyson’s Spheres 😳
There are probably lots of ways to explain this including the multiverse theory but what's clear is that there is no one explanation that makes sense. This paradox has been bugging me way more than Fermi paradox and it still has no answer. Maybe AI will help in the next decade if that competition I mentioned finally hits 100 percent accuracy
@@whatdamath Yeah. We all remember folding@home but I wonder if decentralized simulation has made significant headway in answering this? Regardless, it took Earth a long time and lot of near-deaths to get from photosynthesis to RUclips . . . one hopes that sometimes that journey can happen in 50 million years instead of three billion.
@@whatdamath multiverse seems like a pretty daft way to explain protein folding. It's more likely that real systems don't sample the whole space, at least not in any sequential way as is implied. There is a prior art here: Feynman's approach to quantum mechanics was to have particles take every single possible path to their destination. Feynman was standing on the shoulders of giants, as the calculus of variations had lied the foundation for the idea.
@@whatdamath it appears to me that this is the conflation of two ideas - one about how proteins "know" how to fold the correct way and one about how proteins came to the way they are in the first place. life didnt start on this planet because the rubisco(sp?) protein spontaneously came into existence, but the rubisco protein was the result of many (millions if not billions of?) years of evolution. it reminds me of the watchmaker argument, used in some creationism arguments, of the likelihood of coming across a working pocket watch in the middle of a forest and believing it to have spontaneously form (as a sidenote, i wonder what the number of total organisms that have ever existed, from bacteria to humans, is) how proteins know how to fold is a different problem and i dont know that. but i think that life had much simpler and humbler origins which would limit the complexity by many orders of magnitude and make it much more likely. how many ways of folding does the most simple self-replicating protein have?
Anton, I don't know if my previous comment influenced you to create this video, but I want to say thank you for making it. Your channel as I said before is excellent and you providing more intimate information about you past was honest and heart-felt giving me a much better understanding of why the Fermi Paradox comes up as a major force behind your beliefs. Sometimes gaining a tiny bit of insight into the experiences someone has gone through to become the person they are now is essential to getting a better understanding of their beliefs. I wish that I could meet you in person just to learn more about the experiences you have had in a much less vulnerable environment. Thank you for making this video. I am a long time subsciber and although some of my beliefs don't align with yours and vice versa I know from my perspective that's what makes you interesting to me. Take care and I look forward to your next video.
Some good observations on this so difficult problem. Understanding life (and, worse still, its origin) is a unique scientific problem where intuition from other fields often fails because it is the very rare, very exceptional behaviour which ends up being most important, through the magic of natural selection.
one of the less talked about features of AlphaFold is that it works by taking existing proteins as templates and uses their similarities to predict folding of new sequences. It does not actually work for proteins that are radically new compared to its knowledge base.
That parallels Nature in its making more of what it has already managed to create. It suggests that AlphaFOld embodies some 'shaping' mechanism that can be distilled somehow somewhen.
@@alphalunamare no, AlphaFold uses a database of already analysed proteins (by human minds) and then looks for analogs in new proteins to make a match. Radically new proteins will not be well predicted.
Having designed and synthesized peptides from scratch, there are a few things to consider : pH, ions, ion strength, temp., polarity of solvent. Then, size of the amino acids side chain, charge, surrounding amino acids etc. Amino acids have a preference to form either a alpha helix or beta-sheet, or unordered domain. (See anti bodies). Those domains interact with the others, e.g. a partial hydrophobic alpha helix might connect to another helix (e.g. leucine zipper) or a beta sheet. Etc., etc. Then you have chaperons, post-translatory modifications, etc. A good portion of protein misfold and are recycled. And yes, protein folding prediction is an amazing and difficult task. Machine learning will change the game. A friend’s company is making drugs based on this approach.
Cute. Probably Anton himself is an alien and he is trying to prevent us from discovering the presence of aliens while they steal our patents and technology, a bit like super-Chinese, I would guess.
He is an alien like me, a Canadian in SouthKorea. All of us “ aliens” must have an “ Alien Card” that we must have in us at all times when we go out. They don’t care about your passport but your number on the Alien Card which is in the government computer system. Everything depends on that especially health care. Yeah so I’m an alien and have a card to prove it.
Questioning our beliefs, staying skeptical, is how we get closer to aliens. Thinking we have everything figured out is how we miss them even if they stood right in front of us waving and saying "hey I'm an alien!"
“Big shout out to Anton Petrov! Your channel is a constant source of inspiration for artists like me. The way you present scientific ideas and discoveries, even those that have been around for a while, with such clarity and enthusiasm, is truly remarkable. As a musician and artist, I take the news and insights you share and transform them into creative expressions. Thank you for sparking curiosity and creativity in all of us. Keep up the amazing work!”
Hearing levinthal’s paradox referred to as “an obscure biochemical paradox” is absolutely sending me as someone who does biochemistry. It’s the first “outside looking in” experience I’ve had in a while, thanks.
This is so funny to me, I'm studying Physics and know all about the Fermi paradox but absolutely nothing about biology, the "outside looking in" thing here must be so strong
It is only, literally only, fools and ignoramuses who say, "There is no God." You WILL meet him and you WILL answer to him for your life. Period. The absolute hubris, arrogance, and debased stupidity of the modern man never ceases to amaze. You need to take about 15 seats. Don't be so incredibly mentally deficient that you cannot recognize that higher reasoning and ideas are not results of evolution. One must be objectively idiotic to think evolution has anything at all to do with the formation of reason, logic, and ideas. Educate yourself before making yourself look a fool to the masses who know truth.
@@src3360 Depends on the group. My old pastor told me as a child that science is the method we use to explore how God created the universe and that the harder you look the more intricate the design. 20ish years later I'm no longer religious but I always thought that that pastor was pretty rad
I hadn't heard of it myself but when I read about it, I thought there would be a lot of people saying this. Seems like a fairly big deal for anyone who works on or adjacent to protein science.
I love your videos dude!! I have been watching for a bit more than a year and I want to say your content has helped inspire me to go back to school. Learning awesome facts about our universe made me realize school was very fulfilling to me. I have started taking environmental science and astronomy and have just ended my fifth week. Thanks for inspiring Anton!!!
Wow! The discussion today has been very educational for me! We just do not know enough about life and how it starts to be able to predict how much of it there is in this universe. I am impressed by the fact that science is finding life all over this planet, where it was very much unexpected, such as deep underground and around sea floor vents. Life on earth, at least, proves pretty tenacious, which is encouraging. Hope it translates to life elsewhere! Thanks for another thought provoking segment, Anton!
I try to fill my YR channel subscriptions with presenters who not only know their field well, but also seem to be great people. Anton, you are one of them. Thanks for giving great information, for being so positive, and for being a joyful presence.
Iv always been skeptical over the idea that we need to find a planet exactly like earth to find life. My biggest issue is that earth was extremely different when life started compared to earth now. Chances are life simply adapted to modern earth and we are looking for the wroung conditions. I would also assume with so many possibilities that the proteins on earth may only work on or form on earth. We can't possibly predict what every possibilitie would lead to in every possible condition. If and when we find life i imagine it will have different solutions to similar problems.
I tend to suspect that the "Earth-like planet" thing is for public relations. From a science standpoint, a totally different form of life on a non-Earth-like planet would be more interesting. But the masses who pay the bills want to fantasize they can move away from (or send away) people who annoy them.
We have this perfect world, for our way of being. But don’t appreciate this well enough, to look after our home. If only, we could take a religious approach. To looking after, what we have and rely upon. Buying an EV, enabled by children digging Cobalt, and by burning coal, is not the answer.
Aren't they just searching for planets with amosphere and water, which are prerequisites for life? I've never heard they are searching for exact earth lookalikes.
One thing about the pursuit of scientific knowledge, and really all knowledge, is that none of it is really wasted. Also, it is an endless source of interest. Those who love to learn and are curious will never be bored. There are always new and wonderful things to learn, and times when the things you have learned in the past come together with things you are learning currently, often in unexpected ways. You see that here with Anton, where he is able to draw on his knowledge of biology to consider the possibility of our existence being a rare, if not singular, situation.
Dear Anton, A most interesting video! Thank you for the personal insights you gave at the outset. I’ve watched you for quite some time now, though not always consistently, and I’ve always valued you as a person and someone who has things to say about topics that are important and that also interest me. My heart still breaks for you, for instance, at your loss last year. Yet you have carried on, and that makes you amazing (though the hurt will never leave you). To me, you are a fascinating person. I wish I knew you otherwise apart from your videos. All the best to you. ✅🍂🍃🌈
Isn't the solution to the paradox evolution? The folding is built on combining pieces to larger pieces. So there are some common sections that are stable and can then build up bigger ones.
@@orbismworldbuilding8428 Natural selection? That is a mechanism of evolution. Inorganic matter is indifferent to evolution, and by proxy, natural selection
@@JabberW00kie if these non-living compounds and structures have features which destabilize them or hinder their basic self-replication, then those ones might fail to proliferate. If they have some feature that doesn't hinder significantly, or even benefits that structure's replication and longevity then it might proliferate. Thats natural selection. Also, proteins folding is irrelevant in earths history until you get to the earliest replicators, so the thing about proliferation and replication applies. Things like viruses also have natural selection, and are not alive.
Sorry for your loss Anton! 🙏 Truly grateful for your choice of direction!! Some of the best stuff on YT! You're 1 of the only channels on YT I trust for real intelligent, informative information! You, Anton single handedly sparked my long time love of science and space all over again!! Thank you Anton! You awoke my inner child curiosity, love, passion and excitement for space and science! 👍❤
Anton's foray into the biological sciences is still nascent but i admire his enthusiasm and we can all keep that in mind and offer positive and wonderful information 🤗
In my mind the solution to fermi's paradox is always to assume that we're early, and other life just hasn't had the time to luck out with their proteins. Or that the distances between occurrences of life within space-time are too large or few and far between, making communication / detection impossible.
@@EpicLib A darker solution would entail that once a planet develops intelligence, it will inevitably self destruct before it can colonize its solar system or galaxy. So intelligent* life is not only rare, but very brief. Maybe humanity can overcome this, but I'm having my doubts.
the problem is that's highly unlikely. it's much more likely to be an average civilization by definition. the distances can still be overcome by a single civilization deciding to colonize the galaxy which wouldn't take long on a cosmic time scale, maybe a million years.
@@cripplingautism5785 A million years for flesh and bone is not possible. You'd have to colonize a galaxy with self replicating robots- computers. Data coming back from the Voyagers are grim, showing that interstellar space is full of plasma and electromagnetism- things that would cripple silicone based life or computers. It's just science fiction.
Really glad the comments have people pointing out what was missed with this idea. (The problem isn't how difficult it is to find one complex permutation of one protein, it's how to find and improve on functional proteins at all, which is much much easier. Especially because once you have a single functional organism, you quickly get an exponential increase in experimentation. It's a parallel problem, not a sequential problem. It would take forever to make a single functional protein if you were just trying one after another, but nothing does that. Everything is trying all the time.)
Thanks for making this presentation. This is a good example of just one of the current limits of our knowledge. It brings to mind the words Shakespeare wrote for Hamlet: "There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy."
This is the same « paradox » as saying that the odds of a random sequence of characters being a particular Shakespeare play are astronomically small but languages and proteins aren’t formed that way they’re from successively rearranged small subsequences and building to *some* large body from that is quite tractable even though you’ll almost certainly never come to the same particular one again
The difference being that the works of Shakepeare were the works of Shakespeare and not randomly composed, like seems to be the case with (bio-)chemistry.
@@MartijnHover I think the key word there is "randomly". Maybe there is an over-arching guide to the structure, just like language is the guide to Shakespeare's plays. Just because we don't see it doesn't mean it isn't there. "There are more things on heaven and earth..."
With language and plays we know how intermediate forms are useful and build on each other, like evolution. To know how easy or hard it would be for this protein (or one that functions similarly) to form elsewhere, we need a better understanding of the intermediates and how inevitable the progress was. Until we understand that, we can't say whether life is common and inevitable, or whether we are the only example in the universe. Anton's point is well taken. No matter how many planets are out there, it's not enough to conclude that life should be common until we know how life begins. The Rare Earth Hypothesis is reasonable (maybe most reasonable) until evidence of other life is found or natural processes that lead to life are better understood.
@@MartijnHover protein and nucleic acid sequences are absolutely not randomly composed - in fact it's specifically the fact that they are evolved with point mutations and cut-and-paste from simpler sequences that allows AlphaFold to work ( and it's how we can use molecular biology to infer phylogenetic trees ) On another world the vocabulary would likely be entirely different but it'd similarly be a tiny subset of the high-dimensional potential sequence space
I think what most people aren't getting is the sheer mind-bending level of the numbers in this case. 10^80 is the estimated number of particles of matter in this universe. to get even close to the level of 10^300 that anton is talking about, you'd need to at least cube the number of particles in our universe to get anywhere even close to the neighborhood of the number that he is referring to. To make things even more in perspective, 1 chance in 10^80 is what is considered the limit of what is considered mathematically possible. This number is extraordinarily beyond that. I have an issue with the level of chance here.
Bless you for admitting openly that you had a crisis! That takes introspection and strength. You are stronger because of it. I wish more people could just free associate issues. It is healthy. PEACE!
I'm absolutely no scientist, and yet i love coming back to these videos again and again. You do such a wonderful job at explaining scientific concepts so far advanced and out of my own scope. Keep it up!
A scientist practices the scientific method by definition. Whether you choose to be aware of it or not, we are all scientists by definition or we'd not have lived long enough to reach adulthood. Don't confuse a method with *knowledge* in a niche subject matter. Therefore have more confidence and don't downplay your success :)
@@leudast1215 I think they mean *career scientist* as in someone who has enough of an understanding of a given field to apply the scientific method in a novel way. We all tend to use scientist as a shorthand because most people don’t consider the mundane applications of the scientific method used in everyday life, for good reason: it’s not a useful term
The topic of how likely life is to occur in the universe and to what scale is one of my favorite issues. Throughout history we have seen so many predictions of the likelihood, some that contradict our data looking into the universe and some not. But many of these predictions and their methodologies provide a really interesting snapshot into the scientific understanding of their times. I’ve always wanted to see a larger work in the form of a book or larger article summarizing the history of these predictions and how they’ve changed over time.
It's funny how we've only been looking for a few decades, but we're already trying to explain why we haven't found anything... Idk, it just feels cocky. "Well obviously we'd be good at finding aliens, so the must not exist cause we haven't found any yet." I know that's not how it works, but that's always been the vibe I get from these paradoxes/theories.
Thanks for bringing this engaging topic Anton. One thing that I find kind of puzzling concerns the idea that if we can determine how alpha fold comes up with its predictions it could explain ‘how the process works’ and thus resolve the paradox. But IF alpha fold uses existing, known structures, to predict new ones, that doesn’t seem to suggest it ‘knows’ or can shed any light on how the actual process works, but rather, it only shows that it is good at assessing known patterns and making predictions based on those known patterns. Kind of like large language models are really good at recognizing patterns in text and then predicting correct speech responses, but do not really tell us anything about how we “know” what the speech means because LLMs don’t “know” what they are saying. Am I making any sense?
Isn't that already a good jump forward? just seeing what the AI is assessing or what it groups together could already be good jumping off point. I don't know much about getting this info out of AI, but I've seen atleast one video talking about team doing this successfully. It could basically give a heat map about what the AI was focusing on in a video game, what it saw as negative and what positive (how it would affect it's reward function) That is to say, there is precedent in being able to get info out of AI's black box. Video in question is Robert Miles "We were right! Real Inner Missaligment", and article later in the video about regognizing what AI thinks is "Understanding RL vision" by Distill (found in videos description or first result in google)
@@Trombi01 Yes. Makes sense that the 90% correct results could highlight the similar patterns or sequences that led to the result. These could then be taken up as avenues of research whereby organic chemists may one day be able to uncover causative principles behind the observed phenomena. Well, at least it’s a possibility. But I wonder if maybe we’re rather like the economist in the joke where three men - an engineer, a chemist, and an economist - are stranded on a desert island with only a can of food. As dinner approaches they get hungry and begin discussing how to open the can. The chemist proposes that using a combination of salt water and other substances they may be able get it open through some corrosive process. The engineer suggests dropping the can from a platform (to be designed and constructed) of a given height. Finally, the economist begins his proposed genius solution with the words “let’s assume we have a can opener. Then …” Is the AI route like the economist’s solution, only we actually do already have the can opener? We’ll be able to open many cans, but not without the “already provided” can opener, i.e., the existing complex protein structures, and never actually figuring out origin and inner workings that led to the can opener in the first place.
Protein folding, that brings memories from the early 90 when I was working as a computer tech assistant for post docs who were working on this topic. Those were the days of having a 40k euro Silicon graphics workstation on your desk 😂
I have read papers invoking quantum mechanics to explain how proteins assemble so quickly given the complexity of this problem. They were difficult to understand but implied there is even more going on here than just regular chemistry.
@@jyjjy7 but feynman said that there is not a material reality, only mathmatics exists; i dont believe in that serial sexual offender, i dont believe in his idealism, and i dont believe in quantum physics in general
@@johnro8ot I don't know what you mean by that. The standard model of particle physics, literally the foundation of our understanding of the properties of particles, atoms and molecules, and how they behave and interact, is derived from quantum field theory.
@@jyjjy7 and what I mean is that Newtonian physics, called classical mechanics, accurately describes protein folding. There is a field of science called computational biophysics, look it up, and see what models people ACTUALLY use to simulate protein folding, and stop living in science la la land if you want to actually learn something.
Love the way you used "psilocybin" in the background when talking about medicine ("medical fields"). I believe it has the potential to be a big help to many people suffering from depression and possibly other mental illnesses.
@@Vicki_Benji 'Only the lazy laugh'? What a stupid statement. Where on Earth did you dig that up from? Any more deep insights into human nature - can't wait! 🙄
Maybe photosynthesis is the exceptional odd man out and chemosynthesis is the norm. This could mean that we simply don't know what to look for in way of a life signature... hence the word "alien".
A principle one must consider is that if it can happen, it will. Where there's available energy the potential for some kind of life exists. For all we know, stars are conscious beings.
But even here on earth different types of photosynthesis have evolved. This makes it more likely that it is not so uncommon, if life evolved somewhere else.
@@wolfgangkranek376 If there's 1 to the power of 300 different configurations then it seems there could certainly be more than just that one particular form which could do the same thing. Maybe what we have just happens to be the configuration that worked here 3.8 billion years ago.
You were a great son, remember that. I'll take you on as a second son, perhaps I could even be your grandmother at 77. A Doctor of science, emeritus, at Wake Forest University turned on my love for science, from the micro to the macro--I was getting my Masters in world religions when I went to one of his lectures and decided to take his class on The Universe as an elective. It was mind bending and quite glorious. That was in 1995 when I was 50; I have never since lost that joy of discovery. You have been nourishing me as of late. I keep up with everything in Science and Religion. Continue on
Doesn't the fact that the proteins fold autonomously, into amazing structures, make it more likely that life would be on other planets? Why would protein fold in amazing ways on our planet only?
Answer to your question is in a way in your question, just that the premise in your question is a bit wrong I would say. You ask, "Why would protein fold in amazing ways on our planet only?". Well, this is our planet only because proteins happened to fold here in a certain way. If it would've happened in some other planet, it would be our planet and not this. And if proteins would've not folded in a certain way we wouldn't be here observing it to fold in a certain ways. I hope you get what I'm trying to say here? :)
@@NicholayN There are probably some mechanisms that prohibit many combinations from ever happening. Meaning that the combinations are wayyyy lesss than 10^300 or whatever. Also, maybe, some combinations of all these supposedly 10^300 can give legit functioning proteins with similar properties. There is not way scientists tested 10^300 combinations and only ONE was funtioning. You know what i mean. I see such narrow thinking by the scientists, but what do i know? i am not studying this field.
The resolution to the Levinthal paradox lies in the fact that protein folding is not a random process. Instead, it is guided by the amino acid sequence of the protein and the biophysical properties of amino acids, which drive the protein to fold along specific pathways. These pathways lead the protein to its native, functional structure more quickly than a random search would allow. This concept has been further supported by the development of the energy landscape theory and experimental observations.
I really think the moon and tides was instrumental. Because this means that any proteins and molecules get left high and dry for hours drying out, then wetted, then dried out, then wetted. Oparin and Miller had the right idea, we just need to realize how rare a planet like ours with a GIANT moon, iron core so super duper radiation protection.
With that premise, planets with multiple moons and suns and faster tidal cycles would potentially arrive at viable biochemicals faster than our single sun/moon planet system.
@Fummy Because a moon as large as ours has actually never been found that I am aware of. No planet in our system had anything like the mass ratios earth/ moon does
That's what crossed my mind also. And can life forms only exist if based on proteins? Maybe that requires a whole new theory that we can not even theorize on.
I'm really glad I came across your channel. It's funny enough things that bother me, and when im searching for it and answers, it is rarely understood and the past couple of days your channel came accross that Really elaborate on the things i want to know about.. thanks for the content. You are awesome
Also important is, how close to that exact configuration does the sequence need to be to produce a functional RUBISCO analog. Maybe the gene only assumes the sequence it does because that is the most efficient form, but other alleles may be perfectly functional, just slightly less efficient.
Exactly! Of course the chances of nature folding "the one" Rubisco we know on Earth are insanely low, but who knows how many just as funtional, if not better analogs there are? In proteins, amino acids can be swapped out to some degree and it won't make them unfunctional. I assume that's the case not only with amino acid sequence, but folding as well.
If we truly are alone in the universe, there seems to be an awful waste of space... Plenty of room for humans to trim down our zero sum game behaviour and get out there. Anton, I am glad that you are finding your way and while on your journey you chose to enrich our lives with every post. Very inspiring and humbling. Thank you.
It's still statistically very unlikely that we are alone, none the less. But it may be that the conditions that allow for advanced lifeforms to develop are pretty rare, so maybe other civilizations existed or will exist but we'll never meet them because they are long gone or they still have to develop. It's even possibile that there are civiilizations in the same time frame as us, but so far that we can't really perceive them nor communicate.
actually, space is so big, we can't' even travel in a human life time to the nearest star. Sending humans to Mars and have them survive the trip and doing whatever we want on Mars with very high probability is already a huge task we don't know how to solve yet.
Satan said similar thing. Only we are the waste of space. Just as God gave us dominion over earth,he has given us the entire universe,we are still stuck here in this infernal prison.
I'm so glad you turned up in my program. It's been a while. Most count of how many T-shirts I bought when you were struggling. Just an idea here. We do have a gazillion proteins, but we are still mostly Water. I love water, but I'm very picky about proteins. Plants only, despite something I learned from one of your videos! Keep shining!
Has there ever been an attempt to explain why bacteria grows so rapidly in outer space? Could it have anything to do with a certain kind of radiation? This has always intrigued me and makes me wonder about the concept of panspermia. Interestingly enough, there's even an ancient Egyptian myth about life arriving to earth via asteroids.
@@deltasyn7434 Not really. I just checked and found tis paper: "Effects of spaceflight and simulated microgravity on microbial growth and secondary metabolism" Read it.
@@harmless6813 Sounds like you're trying to be a contrarian. This study confirms that the observation is correct as microgravity is what you have in outer space. I took a wild guess about it having to do with radiation, but I never presented that as fact because I don't know. That said, it seems that lack of gravity and cell motility is the determining factor causing the rapid spread of bacteria in space-like conditions. Which answers my question. Thank you.
@@deltasyn7434 As far as I can tell, there are _some_ bacteria that grow faster in space. Your original made it sound like bacteria do generally grow faster in space, which does not seem to be the case. Sorry if I misunderstood.
I think one solution to the paradox is, that the assumption of a free space around the protein, which would allow for the scanning of all possible configurations, is not true. The ribosome already restricts the free space a lot. Also the precursor amino acids are already having a certain conformation. They do not check out any other conformation, because they are already in their ground state structure. It seems to me similar to diffusion process which dimension is restricted, thus going on much faster than expected for the full dimensional space.
It was always recognised that life is a difficult thing to explain , just because you now have numbers to put it into a framework does not change anything . I love your videos , keep putting them out please.
Thanks! Anton! Your videos are amazing! And some how just listening to you soothes me! Instead of thinking about worldly issues! I focus my energy on gaining knowledge! Your so appreciated ❤
Abiogenisis itself may be a filter to life since we cant actually explain how it began. It's a very interesting topic given the environments the Earth has been through to even allow life to begin at all.
Human-level intelligence might be rare, but persistent complex life (I'm going off the genetic structuring aspect of eucaryotes here) probably leads to "intelligent" life proportionately common. Even a cricket is semi-intelligent, and those are so stupid that they'll eat each other while a tarantula is literally eating them both. The bigger aspect is more likely the amount of possible life, as Earth seems to be on the leading edge of a high-chemical-variety period of the universe: we are among the "early birds" in the universe.
@@absalomdraconis I might speculate the same. And there are man factors, such as the galactic goldilocks zone...our over all soup is richer than previous stellar generations. Cool Worlds speculates on this in fact.
It's only a guess, but I've always thought that life is common. Complex life, probably rarer. Intelligent life, quite rare indeed. I'm not entirely sure that humans count as "intelligent life," but I guess the next three hundred years or so should answer that question.
I have been working on this for many years, in the past, and there were no software tools able to provide any meaningful answer (I am a medicinal chemist, in any case). It seems now we have something working quite well in predicting protein structure. Explaining why, however, is another story. In any case, you don't necessarily need a specific protein to get a specific function. For instance, mammals use hemoglobin to carry oxygen; but some cephalopods use hemocyanine (based on copper instead of iron) to do the same. So I think to focus on a given protein, however relevant it may be to us, is not a good answer. Besides, we use 20 "natural" (to us) aminoacids, but somewhere else they could be more and different. Or may be no protein based life. We are still too much anthropocentric
Mr. Anton, I enjoy listening to your dissertations and there are times when you define the probabilities in your subject in such a way that demands deep thoughts and this time is no exception. Listening to this topic, what you've stated makes me want to apply what seems could be obvious, there was an intervention of sorts? Just an old man pondering the evidence at the scene of of a profound event... Have a good day.
It has been a long time since i studied this sort of thing, but i remember hydrophobic and hydrophilic interactions were great influences in the shapes of proteins, and these are in turn influenced by the properties and sequence of the individual amino acids. these interactions would greatly reduce the degrees of freedom. also the amino acid chains are built up one amino acid at a time so they would start folding as they went along. the rybosomes that assemble the amino acid chains into proteins are incredible "machines", they probably have (its been a while i cant remember for sure!) functions that also facilitate and guide the folding process. also there is certainly a whole host of additional proteins and sugars and nucleoties providing suport in the background. The AI is probably seeing patterns from the effects this additional cellular machinery has on folding, it will probably get better still.
I am wondering two things. First, are we assuming that of a given sequence, there is only one way it can fold for it to be functional? Of the 10^300 different combinations, could many of them be sustainable for life in a given circumstance? Second, is 10^300 the odds for that protein to fold a certain way at the given conditions here on Earth?
I think that number isn't that important. I'd guess the factors that influence WHY it folds is far more important, so the original number is relatively arbitrary.
I bet your mom would be very proud of you and what you have done with your ability to spread knowledge and empower people more than majority of doctors do. Thank you for being you
The other factor contributing to the rare earth hypothesis is time. That is, even if the massively improbable circumstances happen twice, they may happen at completely different epochs.
Maybe other life has other forms of life out there have vastly different base configurations from our own, but if it's equally complex, then it doesn't really answer the question.
Exactly. Everyone says we've not found life anywhere else, but that may or may not be the case. We can only look for life that conforms to the only type of life we know - what exists on this planet.
Those proteins are built one amino acid at a time and don't just pop into existence. I guess their shape is pretty much determined by that sequence. Though this points to a precurser, maybe the RNA world and earlier chemical cycles that led to what we have today
Yeah, that part about deciding in milliseconds the right structure doesn't sound right, apart from chaperones. The sequence will limit the number of configurations possible very strictly and evolution will find a way to fine tune the structure or adapt it to other functions. Abiogenesis remains a mystery, but once you got something that reproduces, even if it's just a simple chemical reaction in a specific environment, it can take off from there.
That is interesting. Do proteins fold while they are assembling one amino acid at a time? If so, it seems that the next one in the sequence would be somewhat nonrandom and the next building block and folding event would happen at the same time following principles of least action or lowest energy or however you'd like to express it. Doesn't the shape at each depend mostly on geometry and whatever lowest energy state applies? Can partly formed proteins somehow join to create bigger and more complex proteins? If this can happen, do they each retain whatever folding they had, or do they unfold and refold?
This wonderful person, is, well, wonderful. Anton, you are one of the very few channels that I can learn something to talk about in conversions with friends and charm them with interesting knowledge. I always though the most important part of the Drake equation was the "fraction of life-supporting planets that develop life". But we only had proof of one planet that could do that, Earth. Until we figure out how life can start, we will never know if it can start elsewhere. Now that you attach a paradox to it, it makes it a lot more credible. In the next party I will say: "have you heard of the Levinthal paradox. No?! Ok, let me tell you."
The way you describe it, it's a wonder protein folding is so fast and consistent even just here on Earth. So many ways to fold a protein . . . and yet they somehow know to fold in the correct manner over and over again. Something must be governing the process that we do not yet understand. Something that allows enormous speed and accuracy where neither should exist at all. Perhaps some influences from the subatomic level that make certain types of folding inevitable and others not.
We do pretty much understand it, and it is way less esoterical than you presume: 1) Amino acids (aas) can be grouped into water attracted and water repelled, so by swimming in their natural surroundig of a cell (water based) the water attracted aas will be attracted to the outside and the water repelled aas will be "clumped" on the inside of the protein. 2) "Chaperone" proteins exist that act like a scaffold or mold to help other proteins fold. 3) Due to 1) and 2) and also other factors, the correct shape of the amino acid is always an energetically stable form. This means folding is less like random movement and more like a ratchet mechanism, single steps can move towards the correct shape, but not towards an incorrect shape since this would require energy and the laws of theromdynamics forbid this.
@@mateolaskowski1731 Thanks for the explanation. Non-expert here, as you may well tell. But it reminds me of the molecular mechanisms of Mad Cow disease I've read about where rogue proteins cause other proteins to fold incorrectly in a pathological manner.
Similar to the situation with growing crystals of new compounds for x-ray diffraction. Nearly impossible until achieved, then inexplicably around the world, all of a sudden much easier.
@@JoeSmith-cy9wj That sounds a little supernatural to me. I'm afraid I'm going to have to invoke the Sagan Standard on this: extraordinary claims require extraordinary evidence. Or does it just get easier because improving lab techniques are copied? Please cite credible scientific sources for your claim. Peer-reviewed if possible, And no flakie social media citations.
@@douglascutler1037 I'm not here to debate anything. For the record, I hate religion, don't believe in god, and do believe in aliens. The crystal thin is just a known fact. How much hearsay and how much recorded evidence I have no idea. But the story has been around and prominent in that field for years. I'm sure Wiki might have something on it. Scientists seem to agree it's odd. I don't know if any statistics or data has been collected.
@@lionelmessisburner7393 the universe has to be infinite. Well but the universe per see because that is just the finite limitation that we are aware of. But beyond that. It's infinity simply because how could something be absolutely finite? That isn't possible.
@@sonkeschmidt2027 I often thought of that too. But it could be something where it loops back on itself. When u get to the end of one side you just find yourself on the complete opposite side. I don’t think that’s the case though, likely it’s infinite
@@lionelmessisburner7393 even if you loop it, it would be infinite. Because the looping would never end. Except how do you know you are looping? In order to know that you must remember that you've been there before. Without that memory you wouldn't even know that you were looping, you would just go on infinitely.
I'm extremely skeptical of the 300 zeros theory. It would mean that things are totally random and yet they aren't. As we all know, we now have AI that can predict quite well how something will fold. This can only mean that such folding is very deterministic, or else AI would output random garbage
Hi Anton, I'm a chemical biologist. There is an entire class of proteins called chaperone proteins which includes heat shock proteins (e.g. HSP60). These proteins assist unfolded proteins to fold the right way to ensure proper tertiary structure. These evolved as proteins got bigger and more complex. I work on an amyloidogenic disease which has dysfunctional chaperone activity as a component. Obviously the first spontaneously formed proteins/enzymes didn't have the luxury of chaperone proteins, but they were likely much smaller and their function much simpler.
The first spontaneously formed proteins are mythological entities and not real things.
This man
You probably just Occam Razored this conundrum.
Do the chaperone proteins embarrass the other proteins by dancing at their prom?
And this chaperone protein folding process doesn’t even need to involve life, presumably.
It may be astronomically unlikely to have a protein with the same exact sequence that folds in the same way, but many amino acids within proteins (depending on the protein) can be changed without affecting the protein's function. This makes the possible number of configurations that can have near identical functionality while having different sequences rather significant.
Scientists seriously hate aliens
@@joshf9074 And that is appalling considering they have yet to even meet any.
The fact that evolution doesn't have an end goal, what we're made of doesn't necessarily mean other life has to be formed similarly, or that proteins and DNA are the most efficient basis for life. We can only base our concept of life on what we see, not what we're likely to find.
And there are also many, many more proteins that have similar enzymatic functionality that have completely different sequences and structures. This really seems like an argument from ignorance.
Finally GODs creation work aknowleged.
Damn Anton, mad respect. I had no idea you were going for your PHD. You are very humble
The man is a superhuman lol, putting out videos everyday while working on a PHD plus all the day to day stuff everyone goes through, and dealing with a tragic loss, he's definitely something special and our world is better for it.
I have such immense respect for him. I can tell he's holding onto so much emotion. He's my favorite youtuber because of his strength and dedication to his passions.
There are three creators on RUclips I might share. He is not just one of them, but by far the most frequent.
Students aspiring to become “Doctors” pursue an MD; medical researchers / professors go the MS or PhD route.
I only clarified the distinction because a lot of academics / doctors get rather (annoyingly) pedantic about their (MANY) abbreviations.
What's his PhD in?
The solution to this paradox is that during the folding process, proteins pass through a series of INTERMEDIATE states that reduce the number of POSSIBLE conformations. These intermediates are like ANCHOR points that help direct the protein to its "CORRECT" structure. In addition, hydrophobic interactions also play an important role in the folding process.
But are the intermediate states deterministic or is there any level of prior information needed to calculate the probability to enter in these states?
@@BernardoSaab Deterministic. Look at those drawings of protein structure. See the long coils? Those are extremely energetically favorable. That alone eliminates most of the other 'possible' configurations.
Agree, additionally conditions (e.g. temp, pH, salinity, redox potential) also influencing the folding process further limiting the possibilities and as such speeding process. In modern organisms there are also other proteins, chaperones, some of which even further limit the possibilities and speed up the process. The growing protein chain is "anchored" in ribosome limiting one end movement, and reducing options. In prebiotic Earth similar role could be played by some mineral material, like particular clay. The refolding (denaturation and folding again) works also in vitro, crystallographers use it many times to obtain protein for their studies - the process is based on e.g. slow redox potential change in particular physico-chemical conditions, it usually takes about 24-48 h and vast majority of the protein is precipitated because of incorrect folding, yet still many molecules gets folded. All the factors mentioned here and by You others folks has tremendous influence on the limiting the "options" that amino acid chain has and speeding the folding process. Calculating influence of all of these factors would be a nightmare, bigger that forecasting weather though.
I think of it like origami - you have to fold creases in your paper at various points in the construction of the model, otherwise the paper won't cooperate with you and it's very difficult to get it to fold correctly. But once you've made the creases, the paper folds into place easily.
Yes. I think this is the same kind of argument that can be used to refute creationists who claim that evolutionary theory cannot possibly be correct because how could a structure so complex as the eye be formed by a series of random mutations all happening in the just the right order. The response is the order of mutations is not random, but rather the first mutation limits the number of possible second mutations, and then the second mutation limits the possible third mutation and so on.
This is a rare case where I disagree with you, Anton. For a start, the problem is predicting the shape given the sequence of proteins, not the unordered bag of them. (That still made it one of the hardest scientific problems we know.)
But the proteins don't then fold randomly, physical forces (like electrostatic attraction) act all along the chain, causing folds to start at various points. As they fold, new parts of the chain come close enough to affect one another to start fresh folds of the already-folded sections. So there's a logarithmic process at work, as well.
Because AlphaFold 2 predicts the structures with nearly the same accuracy as experimental determination, and because proteins do tend to fold predictably in nature, it means the process cannot be a random one, it must be following physical laws that aren't overly sensitive to initial conditions. So in that sense it's similar to predicting the path of a ball rolling down a surface under the force of gravity (just with the parameters of the problem scaled up hugely).
I completely agree.
I would also say the claim that most of the folds "do nothing" is a bit misleading. A differently-folded protein may have no function within the organism that produced it, which is what we might expect. The various chemical components within the organism coevolved, so to speak, and can't be expected to interact meaningfully with just any random particle (outside of something like a part of an immune system that is built to capture/destroy anything it _doesn't_ recognize). However, a "randomly" structured protein can still have chemical interactions with other specific particles, leading to its own complex systems of self-replication and adaptation.
If randomness does play a significant role in how life initially develops, it could be that the biology on this planet is extremely particular. In that case, if we find aliens we would expect their biology to use all different proteins, or perhaps a different mechanism entirely.
Protein folding is really interesting and there is a lot we don't know, but I think claiming that it may be the primary solution to the fermi paradox is somewhat analogous to claiming that non-Chevy vehicles are very unlikely to exist, since a Chevy won't work properly if its ECM (computer) is programmed even slightly differently.
Very well stated.
I'd like to add, as an aside, that there is (probably) no such thing as 'randomness' in the universe, only hidden variables, so small and numerous as to give the appearance of 'randomness' for their incalculability.
The issue I had was with the 10^300 number from Levinthal's Paradox in the first place. Even if protein folding was randomly determined it's very obvious the probabilities of a specific shape change depending on the environment and the paradox doesn't take that into account at all. Of course you are going to get a excessively large number if you assume proteins being created was a random incident, but that is a current debate in origin of life research and there are many who think evolution began before the first nucleotides even. If you assumed all of the cells in a human body assembled randomly then you would get a similarly ridiculous number, but we didn't assemble randomly we evolved over the course of a very long time. It is feasible that the same would be true of proteins.
I also want to add --- protein fitness landscapes, judging by what we've explored of them so far, rather strongly suggest that if you took any protein in nature and folded it an entirely different way that it'd still have *some* chemical function that could possible serve a living thing.
I think this Paradox is built on fundamentally flawed math in numerous ways, not only be not accounting for the fact that some ways of folding have more likelihood than others, but also, that proteins need not be that specific to perform *a* function. This *might* be an issue with certain basal, universal functions one would need, but these functions would not even be needed in protolife, necessarily. (Judging by the likelihood of RNA theory, probably all you need is reproduction of a single macromolecule to kick start life, and the recent experiment that produced autocatalytic RNA suggests that the search space of RNA structures that reproduce themselves is possibly large enough compared to all possible RNA structures that a planet that produces RNA in bulk will inevitably produce one --- I mean, we produced one by producing RNA in much, much smaller bulk... in a single lab.)
I would suggest that at best, this merely suggests that all the possibilities there are in what functions life can achieve is even more open ended... not like it wasn't insanely open ended in the first place.
@@Rampart.X Have you followed the experiments over the years that tried to invalidate the proposed lack of hidden variables in quantum mechanics. They have all failed. There's a lot of evidence for randomness in quantum mechanics, and none for the hidden variable hypothesis any more, as I understand the current state of things.
The only thing still supporting hidden variables is human uneasiness at an intrinsically random Universe.
You just made the channel even more valuable. I love it when multi-disciplinary science comes together!
Anton so many of us are all glad you didn't become a doctor. Me personally you have taught so many things that have lead into learning so many other things in several fields. I have been able to make connections in society and my surroundings that I never would have if I hadn't watched some of you and other RUclips Teachers videos. Seriously man Appreciate you!
Yes. If you ever have doubts about your career path, just give us a shout out and we'll straighten your ass out quick 😊
Appreciate you Anton. Your videos are always a delight. Sometimes it's just nice to hear the "hello wonderful person" and be reminded it's our ability to learn and wonder at the mysteries that often makes us so wonderful.
Hey dude, really sorry to hear about your mom. Losing a parent is always really hard. Though I am happy for you that you were able to use that tragedy to reevaluate what you wanted out of life and chose to pursue what makes YOU happy. Proud of you buddy
Hear hear
Anton, you have become the premier science explainer on YT, perhaps all of the internet. I am constantly impressed by your diligent attention to detail. You represent the very best of the world of digital communication.
Except for this balls up of a video.
If it's predicted that it should take the age of the universe to fold protein yet it only took a few milliseconds then that in itself could mean something intervened or sped up the process or we are just missing something vital to understanding this
Yep, he's become quite the NDT clone... great for telling us what they want us to learn... or to fool us with...
I remember when I was still a Biochem student my professor strictly warned me to not try working on optimizing rubisco or any rubisco research. He said that there have been many who have wasted their whole careers in doing so, and that if evolution could have optimized rubisco it would've found a way to do it by now.
What are job prospects of biochem? How much do you make in terms of salary
Sweet fuck all, I'm a biomedical scientist, and generally the science industry does us dirty. Engineering is a more practical choice (naturally).
not really, not everything in nature is optimized and not everything can be optimized, it is just in a comfortable valley.
I had a friend in grad school who wanted to do this. The hubris…
@@baronka3404 I'd argue that a lot of things in nature _are_ optimized. At least up to some factor of "perfect".
For one, genetic search algorithm in computing basically mimic natural selection + mutations to find optimal solutions to a problem. And they are pretty good at it.
Nature basically had thousands and thousands of years to optimize stuff. Things that were better lived and reproduced more, things that didn't work died off and got culled.
Secondly, and probably most importantly, things in nature are probably optimized using criterion that are unknown/different from what we think of.
Optimizing one thing almost always come with a cost somewhere else.
Thank you Anton. You are very good at simplifying the complicated, and I love that you share a bit about yourself along the way. 🙂
Your mum would be so proud of you Anton - you are sharing info with the masses that is so important - we are glad you decided to change fields! Thanks for the video - I learnt a lot! Love from the UK.
At 6:40 Anton, when you are discussing whether Ribisco could evolve as the same protein molecule on another planet and function the same way elsewhere, it puts me in mind of the two main types of protein molecules that can carry oxygen in blood, and in this sense it is two different elements that are employed to do that job. In most animals it is Iron in Hemoglobin, but in Octupi it is copper. So even here there is at least two known ways on Earth, of carrying oxygen to and waste gases from cells via the vascular system of of animals.
Horseshoes crabs are also copper based, which sadly leads to their poaching for pregnancy tests
This remark appears to be in the same lines as my initial (layman's) thoughts. The combined facts that there are 1E+300 ways to fold, and that if we change one thing in what we know is working, it won't work any longer, does definitely not validate the conclusion that all of the other (1e+300)-1 ways of folding will do nothing to support life.
Also, 1 in 1E+300 seems way, way to large to describe the odds of any conditions that are needed for the functional fold. But I have to admit that this is based on intuition.
Would it be possible for copper to carry enough oxygen for a large animal?
Channichthyidae, sometimes called crocodile icefish have translucent blood, white hearts, and have somehow adapted to live without red blood cells or hemoglobin. Their blood is cloudy with other proteins and macromolecules, some of which help to keep the blood functional in the cold polar waters.
@@Nerthos "which sadly leads to their poaching for pregnancy tests"
Also their blood can be used to detect other things, so the poor things get "farmed" for it.
Proteins don't fold all at once. They are emitted one amino acid at a time from the ribosome. So the first part will get to fold long before the last part gets assembled. This means there are many fewer possibilities than in the all-at-once model. Because it's easy to add more DNA at the end, it also means that the core functionality (of say an enzyme) will tend to be emitted earlier, with the later parts of the protein wrapping around it and providing support structure.
You beat me to it. The "solution space" is thereby very much constrained.
Also, apart from being produced sequentially and folding sequentially along the way, kind of like the silicone worm extruded out of the tip of your cartridge nozzle, or the shavings coming from the tip of your lathe's cutting tool, there is possibly "instrumentation" present in the intracellular environment, helping the protein folds favour particular macro-shapes. Someone's just received a thousand upvotes for mentioning the Chaperone proteins... voila, some of it is known and has a name :-)
Wondering "why does it fold exactly this way, if it's just a pile of amino-acids, with a 10^-30 probability" is a bit like asking "how does a complex piece of software happen, say the MS Windows, consisting of some 30 letters of the ASCII character set".
Except that here, unless you prefer to believe in some $DEITY, there is no genius creator involved, writing the code. The cellular protein factory and the DNA have co-evolved for a long time, driven by the ultimate in evolutionary genetic algorithms. Over billions of years, and possibly billions of generations, fitness for survival has been the ultimate "scoring function" = the judging force over what stays in and what goes out, of the genetic sequences that end up in the folding proteins being produced. This is how the endless space gets searched for the viable paths.
Yes. Thank you. I only learnt this recently and it makes AlphaFold far less impressive. Impressive it still is though. I think DeepMind would prefer you think there are really 1e+300 possibilities
@@xrysf03 and mr/Dr landman, thank you for these replies.
Your right, however when a protein is melted, via heat blasting it's quaternary, tertiary, and even secondary structure apart, very often they do properly reform when they are cooled down. Certainly not always, but often.
adding to that, there are also so called chaperones, proteines that assist in the folding process of many proteins. Not to mention that amino acid chains are not like a rope which you can fold any way you want. The side chains of the various amino acids have different properties. Some will attract each other, while others repel each other. Then there are restrictions to how much you can twist a chemical bond. At least he mentioned things like ph and temperature affecting how proteins fold. It's all about the forces that act upon each other. Those forced attract or repel leading to the final structure. So there are many restrictions in place, which greatly reduces the number of possible ways a protein can fold. Proteins can fold in many ways, but in the given conditions and the way they are build, there is only that one end result.
And the existence of the chaperones is just a result of evolution. There is nothing paradoxical about it, what works works. If it works it can survive and pass on. If not, then that's that, it will not survive or at the very least no procreate. Life doesn't have to be perfect. It isn't. It just has to work somehow.
Love your channel Anton. I never realized that you lost your mom eventually decided not to become a doctor but do remember you had background in biochemistry. I had sort of the reverse situation in a way. My dad died of ALS when I was younger, so I had decided to become a doctor, but eventually I realized this was not a pure motivation of mine, but was fueled by revenge and my inability to accept death. Eventually I decided to become a math major when I entered college, and rediscovered that childlike wonder in it, unclouded by grief.
Yes....pursue whatever created that child like wonder, that giddiness....whatever that field is for you, you can not go wrong ! Its anatomy/physiology for me. I feel like I'm entering a magical cosmos and I need to understand all of it, and it makes me happy. (Then I can't shut up about it to everyone I'm close with. 😅) Stem fields unite!
I enjoyed blowing up things as a child. I now work on 155 howitzers
I am glad you chose another path, not just because I get videos out of it, but because I have seen many people following a path set by their parents at their own detriment.
If I remember correctly one of the leading hypotheses out there for how proteins fold so quickly is the energy landscape funnel model where the protein initially samples a wide range of conformations but as soon as it becomes slightly more folded the conformations it can sample become drastically limited, kind of like cutting off entire branches of a decision tree. This is in line with our understanding of thermodynamics with a highly flexible state being an unstable one for most proteins (now intrinsically disordered proteins are a whole other story) and the experimental evidence of protein (un)folding in domains and through specific stages. The funnel hypothesis also explains why some proteins are prone to aggregation and how refolding a protein is possible but needs an external boost like a chaperone protein. Great video though, love your content!
Exacrly... proteins do not need to try all positions. They minimise their energy. The same as a marble that rolls down a hill... there are infinite many ways, but the marble will follow just one (at least if it is a macroscopic marble). The trick of living nature is, to control this process by shaping the landscape (through the environment like pH, ion concentration, fatty membranes in the vicinity and the process itself with varying synthesis soeed and chaperone helpers around), so that the marble roles down a useful path and arrives at a designated target point and not elsewhere. When a protein gets unfolded and refolds on its own, then the way down of the marble is more random (regarding the final position) - or maybe not random, but not useful anymore. That's when protein turn nonfunctional. It depends on the sequence.
So Levinthal's paradox is none... no superpower or superintelligence required, just the laws of physics and chemistry. And a lot of trying and improvement over time to get there where nature is right now.
Cheers
Sounds like as soon as it starts moving, it accelerates exponentially to its final state.
pushed by the surrounding environment, hydrophobic collapse (hydrophobic parts of the chain flip inwards and are shielded from the water solvent by more hydrophillic parts) is one of the proposed mechanisms for eliminating a lot of "wiggle room" very quickly, if I remember correctly. I could also imagine that there is some form of folding assist in cells where enzymes catalyse conformation changes but also induce conformation in the first place.
This makes a lot of sense to me, depending on conditions around a protein chain, all the astronomically large number of final configurations it can reach are possible but they are not all equally likely. It's likely a continuous process where at every new step along the way, there is a sampling function evaluating the current conditions and a folding function directing the next step in configuration based on a more limited set of possibilities. It also sounds pretty close to how I imagine alphafold comes up with these predictions although I don't think that accounts for a continuous environmental evaluation at every (infinitesimal) step along the folding process.
may be the deepmind follows similar constraints to its solution
Although machine learning models like AlphaFold aren't mechanistic, they can still be useful for understanding how folding works. For example, because the input for AlphaFold is the amino acid sequence, its success at predicting folding structure implies that the sequence encodes the relevant information needed for folding. Advances in neural network interpretability may also provide insight into what data correlations the algorithms are leveraging to make predictions. Very cool!
I think it rather implies the opposite - that there is extrinsic information (which we don't know yet) not encoded in the sequence but causing the recognizable patterns. If you thoroughly analyzed the entire U.S. census data minus 1 name and then were handed a pile of letters, "nhismjh to", you could probably predict with good odds that "john smith" is a likely name. At least more likely than "otihshmnj ". But that's not encoded in the letters themselves. It's extrinsic things in our language and history and how we form names. What's really cool is that, for things like this (protein folding) we now are developing AI that can do the analysis, recognize the patterns, and predict the outcomes. It can't yet tell us what the extrinsic causes are or why, but it's still a huge step! Could dramatically reduce the area we need to search to find the answers. Exciting stuff!
@@rascta Yeah, that makes sense. I also don't think it's a matter of the sequence encoding anything as far as a meta instruction or something; I think it's a matter of only these specific arrangements actually occur in our conditions here. So the ML is essentially ruling out all of these "possible" arrangements that never actually are seen in nature (at least in our conditions) simply by having never observed them and thus not having them as options to arrange as it predicts the outcome. Those conditions include the ph, temp, etc like Anton mentioned, but also the other proteins that are interacting with it during folding, and other catalysts and potential limiting factors we aren't aware of yet. So we have a lot of fundamental physics happening, which leads to chemical paths of least resistance as bonds try to form in the most energetically favorable arrangement, but that is counter balanced by the direct forces imposed by the local environment (i.e. chaperone proteins), and this leads only a tiny subset of useful proteins ever sticking around and not being reabsorbed by other processes.
Would imagine that the AI is doing pattern recognition. It is recognizing combinations and sequences and eventually we will make another AI to tell us what the other AI did.
@@leighz1962 That is the idea behind RLHF
Folding is SEQUENTIAL. Proteins are produced ONE amino acid monomer at a time, hanging from the assembly point of the ribosome. The first monomer has nothing to fold against until the second monomer is extruded. Now there are two monomers flopping around, but one of them is still stuck to the ribosome. Of all of the possible ways they can flop together, one of these ways has the maximum stability (lowest energy), so the monomers stick together in this most stable configuration and tend to STAY that way. Next, monomer THREE is extruded and comes into contact with the grouping of the previous TWO in the prior maximally stable configuration. So the new monomer is bumped into by the TWO previous monomers, already folded together, which bounce around and form the maximally stable configuration CONSISTENT with the previous folding. There are NO FREE CHOICES, because, at each stage, only one configuration is maximally stable GIVEN the configuration in which the prior monomers have folded. Hence, at each stage, there is only ONE configuration possible. Hence, THERMODYNAMICS drives the folding process. It does not matter if other configurations are possible, even if some of them are more stable than the current configuration, because the previous configuration is already locked into place. I am ignoring the complexities introduced by chaperone proteins, which are already out there.
Good thing to recall a few principles of chemistry.
One slight extension of this theory: all possible minimas are expected to occur, just at different probabilities, there is not one choice only, but many, with one being the preferred one in most cases. This still explodes in combinations, but vastly reduces the space and especially makes for a much more compact space of probabilities.
Thank you. And such rules apply throughout the universe wherever a cosy, long-lived and relatively stable planet orbits. Brings down the odds against rarity.
The eukaryotic jump does seem a rarity, but that’s another problem. Should be more variants.
1. Why are you using all caps?
2. Does it even matter? Why would we think any sequence we identify has to be some kind of unique formula for success toward intelligent life? No biology degree necessary to realize we can’t begin to conclude there’s only one way for it to happen.
Actually, when it comes to polypeptides over a certain length, the kinetics of the cellular environment interfere with proper folding, thus the necessity for isolation inside the chaperone capsule. The nascent protein is literally transported into the chaperone where it achieves its tertiary structure. Even so, the kinetics of such a long chain allow for multiple bonding configurations separated by miniscule energy levels. The Levinthal paradox still remains poorly understood.
Thank you Anton for keeping my brain active and engaged at my age. Although I could never do or understand the science/math behind your videos, your explanations are enough for me to get the concepts you are trying to convey. You bring a sense of wonder and awe to my existence through science.
Seems that way too for a knucklehead that spent too much time with the kids in math for the gifted and talented getting them to work together to write code that could have been created by this that could not, alone. So fun. Good times.
I think it shows good character that you chose to take your own path out of the tragedy. You are educating thousands of people including me, so thank you for what you do. Your ability to make greatness out of life in the face of loss is inspiring to me.
Hi Anton, I am a 58-year-old chasing his Ph.D. At present, I have two master's degrees and am waiting for word from the programs I chose. I wanted to encourage you. Chasing a Ph.D. you want may turn out to be the exact opposite of your previous Ph.D. experience. You are very talented. I have often thought of you in my own removed storyline about you as an astrophysicist or an astrobiologist. I hope you can turn all your efforts into an amazingly self-fulling path! Best wishes!!!!! Rob
This presentation is mind-boggling, as well as the comments it generated. Good work Anton!
We do have a pretty good idea of how most of this is done though : chaperone proteins are shaping proteins as they are synthesized. So while proteins will by themselves fold into a stable configuration, they are also shaped by the proteins they are synthesized around (which is determined by a sequence in the mRNA, which will lead the mRNA to be transported in a particular place in the cell, which may vary with epigenetics). This makes things a bit more complex than just knowing the sequence of the protein to determine the shape, the context matters. Knowing the sequence of the mRNA and its epigenetics, on top of the whole machinery of the cell, is what is really necessary to determine the shape of the protein.
The road map and/or definitions of the configuration to be applied before new protein is made is what I believe his point was.
The video isn't talking about the folding of proteins in a cell, it's the formation of said proteins in nature before any life exists. Chaperone proteins don't answer the question they're a part of the question since they must ALSO form spontaneously at some point for any other proteins to use them as a guide. How did the exact proteins needed to CREATE life form on their own randomly and why not anywhere else......not how does life on earth create the proteins it's been creating for eons. Two different questions.
@@MrBottlecapBill I guess we can only assume, random effects eventually create something complex that carries on being more complex. Pretty strange effect.
@@MrBottlecapBill That's my point : while the very first proteins did form spontaneously and folded uniquely based on electrochemical properties, everything that formed after formed in an environment with preexisting proteins interacting with the formation. Any protein that is common to all life on earth comes from dna being translated, not from spontaneous formation. So the mechanisms have had the time to evolve quite a bit and are nothing like the primordial soup. The environment matters a lot.
(as for why nowhere else : we have no proof life never emerged on our closest neighbours in the first place, so let's not jump to conclusions : for a while there was no life on Earth either, and at some point it will disappear. Time and the lack of data may be the main reasons why we didn't find extraterrestrial life yet, not even talking about a civilization like ours which is going to burn all of the resources of the planet in a few centuries. That would be looking for a needle in a haystack while not being able to feel your extremities and being blindfolded, that's about how bad we are)
@@MrBottlecapBill Assuming a random event on the nature where the protein doesn't even can exist doesn't sound very intelligent... the video states a random chance, who said the folding is truly random? have they acknowledge EVERY interaction? atoms and molecules have physics and chemical laws to obey, it's never random if you scrutinize enough.
I think this paradox can be slightly resolved by recognizing that there may be many potential pathways that life can use for a given function. So initial production of a specific protein might be highly improbable, but reproducing a protein that achieves a similar function could be much more likely.
Yeah but if we can't measure and quantify and then hook it up to a meter we have a hard time understanding or imagining any other possibilities
==== Even getting two proteins to work together is beyond explaining. And ten is worse still. No way the folding protein process originated from dirt. Please explain how this could be? ==== Evolution = Self Assembling Atoms = Impossible. ===
Imagine that each environment where life is POSSIBLE (which we still don't know all of them, because we keep finding life in more and more extreme environments) has it's own set of one, two, (three, thirty, six thousand, or more?) "shapes" which a protein CAN fold into which would support life in that specific environment.
To make it easier (than trying to think about all the huge numbers of different ways proteins can fold) imagine that these shapes are a 2D silhouette for combinations of Legos which is cut out from a board. You have a box of random Legos provided by the environment (the environment favoring some Lego blocks more than others, as per it's unique environment).
Your task (and the task of about a bazillion others at the same time) is then to find just ONE combination of constructed Legos which can EXACTLY pass thru the cut-out shape. Of course, there are many different combinations which will likely work... but you only NEED to find ONE.
Now let's go back to that huge number of 10^300... That's how many different ways the entire set of Legos CAN be put together. But you don't need to go thru each of them one at a time... because biology and chemistry almost never do things like that in series; it works in parallel, so it can try many, many 'shapes' at the same time. In fact, it also favors things that are easier to try and going toward more complex as the assembly continues on.
So you have an environment which favors a subset of proteins, which already cuts down on the number of variations... And you have more than a single attempt at any given time (assuming these amino acids are in an environment where protein assembly is possible - aka: "where life is POSSIBLE" )... Instead you have bazillions of iterations being tried at the same time.
Suddenly this "paradox" is looking like much less of a paradox, yes?
(I think I'll post this in the main reply thread too)
@@JeskaDax Exactly my point!
Something I wanna add is that not all 10^300 permutations are equally probable. For example, hydrophobic amino acids are more likely to interact with each other than hydrophilic ones (and vice versa). This discards a lot of permutations that simply aren't as stable or possible to form in the first place. Therefore, intermolecular forces such as Van der Waals or hydrogen bonds may induce/promote specific permutations. So, is possible that a lot of unstable states collapse into more stable ones. In general, the system will keep collapsing to more stable states which leads to the desired folding state.
This is what i read the comments for
@spacekoala2887
Water certainly must have a major effect on protein folding. But the cytoplasma in a cell contains somuch more, which also should affect it.
I guess that chaperon proreins doesn't even has to be for only that purpose.
However, it is surprising how different initial conditions at the beginning of protein folding result in identical final folded structure.
What do you mean by desired folding state? Weren't these supposed to be just randomly forming proteins?
@@kaufmanat1 No, it's not random.
When I am feeling down, I can always come to Anton's channel and be addressed as "wonderful person"... Thanks, Man...
Directing the infinite permutations of protein folding into adaptive structures might be the actual intelligence behind intelligent life - not iPhones and Dyson’s Spheres 😳
There are probably lots of ways to explain this including the multiverse theory but what's clear is that there is no one explanation that makes sense. This paradox has been bugging me way more than Fermi paradox and it still has no answer. Maybe AI will help in the next decade if that competition I mentioned finally hits 100 percent accuracy
@@whatdamath Yeah. We all remember folding@home but I wonder if decentralized simulation has made significant headway in answering this? Regardless, it took Earth a long time and lot of near-deaths to get from photosynthesis to RUclips . . . one hopes that sometimes that journey can happen in 50 million years instead of three billion.
@@whatdamath multiverse seems like a pretty daft way to explain protein folding. It's more likely that real systems don't sample the whole space, at least not in any sequential way as is implied. There is a prior art here: Feynman's approach to quantum mechanics was to have particles take every single possible path to their destination. Feynman was standing on the shoulders of giants, as the calculus of variations had lied the foundation for the idea.
@@whatdamath it appears to me that this is the conflation of two ideas - one about how proteins "know" how to fold the correct way and one about how proteins came to the way they are in the first place. life didnt start on this planet because the rubisco(sp?) protein spontaneously came into existence, but the rubisco protein was the result of many (millions if not billions of?) years of evolution. it reminds me of the watchmaker argument, used in some creationism arguments, of the likelihood of coming across a working pocket watch in the middle of a forest and believing it to have spontaneously form (as a sidenote, i wonder what the number of total organisms that have ever existed, from bacteria to humans, is)
how proteins know how to fold is a different problem and i dont know that. but i think that life had much simpler and humbler origins which would limit the complexity by many orders of magnitude and make it much more likely. how many ways of folding does the most simple self-replicating protein have?
I don't know about the intelligence part, but it would certainly explain a lot about the adaptability of the evolutionary process.
Folding at home is an interesting program to be able to participate in this research and learn about protein folding too
Anton, I don't know if my previous comment influenced you to create this video, but I want to say thank you for making it. Your channel as I said before is excellent and you providing more intimate information about you past was honest and heart-felt giving me a much better understanding of why the Fermi Paradox comes up as a major force behind your beliefs. Sometimes gaining a tiny bit of insight into the experiences someone has gone through to become the person they are now is essential to getting a better understanding of their beliefs. I wish that I could meet you in person just to learn more about the experiences you have had in a much less vulnerable environment. Thank you for making this video. I am a long time subsciber and although some of my beliefs don't align with yours and vice versa I know from my perspective that's what makes you interesting to me. Take care and I look forward to your next video.
Some good observations on this so difficult problem. Understanding life (and, worse still, its origin) is a unique scientific problem where intuition from other fields often fails because it is the very rare, very exceptional behaviour which ends up being most important, through the magic of natural selection.
one of the less talked about features of AlphaFold is that it works by taking existing proteins as templates and uses their similarities to predict folding of new sequences. It does not actually work for proteins that are radically new compared to its knowledge base.
That parallels Nature in its making more of what it has already managed to create. It suggests that AlphaFOld embodies some 'shaping' mechanism that can be distilled somehow somewhen.
@@alphalunamare no, AlphaFold uses a database of already analysed proteins (by human minds) and then looks for analogs in new proteins to make a match. Radically new proteins will not be well predicted.
Having designed and synthesized peptides from scratch, there are a few things to consider : pH, ions, ion strength, temp., polarity of solvent. Then, size of the amino acids side chain, charge, surrounding amino acids etc. Amino acids have a preference to form either a alpha helix or beta-sheet, or unordered domain. (See anti bodies). Those domains interact with the others, e.g. a partial hydrophobic alpha helix might connect to another helix (e.g. leucine zipper) or a beta sheet. Etc., etc.
Then you have chaperons, post-translatory modifications, etc. A good portion of protein misfold and are recycled.
And yes, protein folding prediction is an amazing and difficult task. Machine learning will change the game. A friend’s company is making drugs based on this approach.
Anton, the aliens appreciate your continued efforts to conceal their existence.
Cute. Probably Anton himself is an alien and he is trying to prevent us from discovering the presence of aliens while they steal our patents and technology, a bit like super-Chinese, I would guess.
That's something an alien would say
He is an alien like me, a Canadian in SouthKorea. All of us “ aliens” must have an “ Alien Card” that we must have in us at all times when we go out. They don’t care about your passport but your number on the Alien Card which is in the government computer system. Everything depends on that especially health care. Yeah so I’m an alien and have a card to prove it.
Questioning our beliefs, staying skeptical, is how we get closer to aliens. Thinking we have everything figured out is how we miss them even if they stood right in front of us waving and saying "hey I'm an alien!"
@@doremysheep7864 No you.
“Big shout out to Anton Petrov! Your channel is a constant source of inspiration for artists like me. The way you present scientific ideas and discoveries, even those that have been around for a while, with such clarity and enthusiasm, is truly remarkable. As a musician and artist, I take the news and insights you share and transform them into creative expressions. Thank you for sparking curiosity and creativity in all of us. Keep up the amazing work!”
Hearing levinthal’s paradox referred to as “an obscure biochemical paradox” is absolutely sending me as someone who does biochemistry. It’s the first “outside looking in” experience I’ve had in a while, thanks.
This is so funny to me, I'm studying Physics and know all about the Fermi paradox but absolutely nothing about biology, the "outside looking in" thing here must be so strong
@Nathan Melia
Very true!!!
As a previously indoctrinated person, god is who creates life. No wiggle room for interpretation.
It is only, literally only, fools and ignoramuses who say, "There is no God." You WILL meet him and you WILL answer to him for your life. Period. The absolute hubris, arrogance, and debased stupidity of the modern man never ceases to amaze. You need to take about 15 seats. Don't be so incredibly mentally deficient that you cannot recognize that higher reasoning and ideas are not results of evolution. One must be objectively idiotic to think evolution has anything at all to do with the formation of reason, logic, and ideas. Educate yourself before making yourself look a fool to the masses who know truth.
@@src3360 Depends on the group. My old pastor told me as a child that science is the method we use to explore how God created the universe and that the harder you look the more intricate the design.
20ish years later I'm no longer religious but I always thought that that pastor was pretty rad
I hadn't heard of it myself but when I read about it, I thought there would be a lot of people saying this. Seems like a fairly big deal for anyone who works on or adjacent to protein science.
I love your videos dude!! I have been watching for a bit more than a year and I want to say your content has helped inspire me to go back to school. Learning awesome facts about our universe made me realize school was very fulfilling to me. I have started taking environmental science and astronomy and have just ended my fifth week. Thanks for inspiring Anton!!!
Wow! The discussion today has been very educational for me! We just do not know enough about life and how it starts to be able to predict how much of it there is in this universe.
I am impressed by the fact that science is finding life all over this planet, where it was very much unexpected, such as deep underground and around sea floor vents. Life on earth, at least, proves pretty tenacious, which is encouraging. Hope it translates to life elsewhere!
Thanks for another thought provoking segment, Anton!
@Anton, we all benefit so much from your analysis and insights, and I personally am very grateful for your regular videos. Thank you.
I try to fill my YR channel subscriptions with presenters who not only know their field well, but also seem to be great people. Anton, you are one of them. Thanks for giving great information, for being so positive, and for being a joyful presence.
Iv always been skeptical over the idea that we need to find a planet exactly like earth to find life. My biggest issue is that earth was extremely different when life started compared to earth now. Chances are life simply adapted to modern earth and we are looking for the wroung conditions. I would also assume with so many possibilities that the proteins on earth may only work on or form on earth. We can't possibly predict what every possibilitie would lead to in every possible condition. If and when we find life i imagine it will have different solutions to similar problems.
I tend to suspect that the "Earth-like planet" thing is for public relations. From a science standpoint, a totally different form of life on a non-Earth-like planet would be more interesting. But the masses who pay the bills want to fantasize they can move away from (or send away) people who annoy them.
@@stevenscott2136 That's the most poignant thing I've heard in a long while.
We have this perfect world, for our way of being.
But don’t appreciate this well enough, to look after our home.
If only, we could take a religious approach. To looking after, what we have and rely upon.
Buying an EV, enabled by children digging Cobalt, and by burning coal, is not the answer.
Aren't they just searching for planets with amosphere and water, which are prerequisites for life? I've never heard they are searching for exact earth lookalikes.
Was it? Life could only form after earth cooled down. If a similar planet is the same temperature now, it was a similar temperature back in time.
I always found it interesting how shapes can create something like photosynthesis. It just blows my mind.
It's not just shapes. There are also forces involved.
One thing about the pursuit of scientific knowledge, and really all knowledge, is that none of it is really wasted. Also, it is an endless source of interest. Those who love to learn and are curious will never be bored. There are always new and wonderful things to learn, and times when the things you have learned in the past come together with things you are learning currently, often in unexpected ways. You see that here with Anton, where he is able to draw on his knowledge of biology to consider the possibility of our existence being a rare, if not singular, situation.
Dear Anton,
A most interesting video! Thank you for the personal insights you gave at the outset. I’ve watched you for quite some time now, though not always consistently, and I’ve always valued you as a person and someone who has things to say about topics that are important and that also interest me. My heart still breaks for you, for instance, at your loss last year. Yet you have carried on, and that makes you amazing (though the hurt will never leave you). To me, you are a fascinating person. I wish I knew you otherwise apart from your videos. All the best to you. ✅🍂🍃🌈
Isn't the solution to the paradox evolution? The folding is built on combining pieces to larger pieces. So there are some common sections that are stable and can then build up bigger ones.
You might be thinking in the right direction, but something more stable doesn't necessarily translate to more functional.
The issue is that evolution cannot be invoked with abiogenesis. Life dances to evolution, but evolution holds no sway over non-living material.
@@JabberW00kie natural selection and structural impurities, however do
@@orbismworldbuilding8428 Natural selection? That is a mechanism of evolution. Inorganic matter is indifferent to evolution, and by proxy, natural selection
@@JabberW00kie if these non-living compounds and structures have features which destabilize them or hinder their basic self-replication, then those ones might fail to proliferate. If they have some feature that doesn't hinder significantly, or even benefits that structure's replication and longevity then it might proliferate.
Thats natural selection.
Also, proteins folding is irrelevant in earths history until you get to the earliest replicators, so the thing about proliferation and replication applies.
Things like viruses also have natural selection, and are not alive.
Sorry for your loss Anton! 🙏
Truly grateful for your choice of direction!!
Some of the best stuff on YT!
You're 1 of the only channels on YT I trust for real intelligent, informative information!
You, Anton single handedly sparked my long time love of science and space all over again!!
Thank you Anton!
You awoke my inner child curiosity, love, passion and excitement for space and science! 👍❤
Anton's foray into the biological sciences is still nascent but i admire his enthusiasm and we can all keep that in mind and offer positive and wonderful information 🤗
In my mind the solution to fermi's paradox is always to assume that we're early, and other life just hasn't had the time to luck out with their proteins. Or that the distances between occurrences of life within space-time are too large or few and far between, making communication / detection impossible.
Well that's an interesting point, the second part you mentioned about time distances.
@@EpicLib A darker solution would entail that once a planet develops intelligence, it will inevitably self destruct before it can colonize its solar system or galaxy. So intelligent* life is not only rare, but very brief. Maybe humanity can overcome this, but I'm having my doubts.
the problem is that's highly unlikely. it's much more likely to be an average civilization by definition.
the distances can still be overcome by a single civilization deciding to colonize the galaxy which wouldn't take long on a cosmic time scale, maybe a million years.
@@cripplingautism5785 A million years for flesh and bone is not possible. You'd have to colonize a galaxy with self replicating robots- computers. Data coming back from the Voyagers are grim, showing that interstellar space is full of plasma and electromagnetism- things that would cripple silicone based life or computers. It's just science fiction.
But we know we are not early.
The most professional RUclipsr out there. Keep up the good work, Anton!
Thanks Anton for this thought provoking video. As always, you manage to inform as well as pique our curiosity for more.
Really glad the comments have people pointing out what was missed with this idea. (The problem isn't how difficult it is to find one complex permutation of one protein, it's how to find and improve on functional proteins at all, which is much much easier. Especially because once you have a single functional organism, you quickly get an exponential increase in experimentation. It's a parallel problem, not a sequential problem. It would take forever to make a single functional protein if you were just trying one after another, but nothing does that. Everything is trying all the time.)
Thanks for making this presentation. This is a good example of just one of the current limits of our knowledge. It brings to mind the words Shakespeare wrote for Hamlet: "There are more things in heaven and earth, Horatio, than are dreamt of in your philosophy."
Anton, I would seriously enjoy being your friend in real life. Thank you for your content and awesome science communication skills!
This is the same « paradox » as saying that the odds of a random sequence of characters being a particular Shakespeare play are astronomically small but languages and proteins aren’t formed that way they’re from successively rearranged small subsequences and building to *some* large body from that is quite tractable even though you’ll almost certainly never come to the same particular one again
The difference being that the works of Shakepeare were the works of Shakespeare and not randomly composed, like seems to be the case with (bio-)chemistry.
@@MartijnHover I think the key word there is "randomly". Maybe there is an over-arching guide to the structure, just like language is the guide to Shakespeare's plays. Just because we don't see it doesn't mean it isn't there. "There are more things on heaven and earth..."
With language and plays we know how intermediate forms are useful and build on each other, like evolution. To know how easy or hard it would be for this protein (or one that functions similarly) to form elsewhere, we need a better understanding of the intermediates and how inevitable the progress was.
Until we understand that, we can't say whether life is common and inevitable, or whether we are the only example in the universe.
Anton's point is well taken. No matter how many planets are out there, it's not enough to conclude that life should be common until we know how life begins. The Rare Earth Hypothesis is reasonable (maybe most reasonable) until evidence of other life is found or natural processes that lead to life are better understood.
@@MartijnHover protein and nucleic acid sequences are absolutely not randomly composed - in fact it's specifically the fact that they are evolved with point mutations and cut-and-paste from simpler sequences that allows AlphaFold to work ( and it's how we can use molecular biology to infer phylogenetic trees )
On another world the vocabulary would likely be entirely different but it'd similarly be a tiny subset of the high-dimensional potential sequence space
I think what most people aren't getting is the sheer mind-bending level of the numbers in this case. 10^80 is the estimated number of particles of matter in this universe. to get even close to the level of 10^300 that anton is talking about, you'd need to at least cube the number of particles in our universe to get anywhere even close to the neighborhood of the number that he is referring to. To make things even more in perspective, 1 chance in 10^80 is what is considered the limit of what is considered mathematically possible. This number is extraordinarily beyond that. I have an issue with the level of chance here.
Bless you for admitting openly that you had a crisis! That takes introspection and strength. You are stronger because of it. I wish more people could just free associate issues. It is healthy. PEACE!
I'm absolutely no scientist, and yet i love coming back to these videos again and again. You do such a wonderful job at explaining scientific concepts so far advanced and out of my own scope. Keep it up!
A scientist practices the scientific method by definition. Whether you choose to be aware of it or not, we are all scientists by definition or we'd not have lived long enough to reach adulthood. Don't confuse a method with *knowledge* in a niche subject matter. Therefore have more confidence and don't downplay your success :)
@@leudast1215 I think they mean *career scientist* as in someone who has enough of an understanding of a given field to apply the scientific method in a novel way. We all tend to use scientist as a shorthand because most people don’t consider the mundane applications of the scientific method used in everyday life, for good reason: it’s not a useful term
The topic of how likely life is to occur in the universe and to what scale is one of my favorite issues. Throughout history we have seen so many predictions of the likelihood, some that contradict our data looking into the universe and some not. But many of these predictions and their methodologies provide a really interesting snapshot into the scientific understanding of their times. I’ve always wanted to see a larger work in the form of a book or larger article summarizing the history of these predictions and how they’ve changed over time.
True! I remember reading in a science magazine as a child (I’m 42 now) that it was probably extremely rare that stars have planets… 😂
@@pipopipo6477 ... until they were found to be in almost every star system examined ... many, in the Goldilocks zone...
@@davidconner-shover51 exactly!
It's funny how we've only been looking for a few decades, but we're already trying to explain why we haven't found anything...
Idk, it just feels cocky. "Well obviously we'd be good at finding aliens, so the must not exist cause we haven't found any yet."
I know that's not how it works, but that's always been the vibe I get from these paradoxes/theories.
Thanks for bringing this engaging topic Anton. One thing that I find kind of puzzling concerns the idea that if we can determine how alpha fold comes up with its predictions it could explain ‘how the process works’ and thus resolve the paradox. But IF alpha fold uses existing, known structures, to predict new ones, that doesn’t seem to suggest it ‘knows’ or can shed any light on how the actual process works, but rather, it only shows that it is good at assessing known patterns and making predictions based on those known patterns. Kind of like large language models are really good at recognizing patterns in text and then predicting correct speech responses, but do not really tell us anything about how we “know” what the speech means because LLMs don’t “know” what they are saying. Am I making any sense?
Isn't that already a good jump forward? just seeing what the AI is assessing or what it groups together could already be good jumping off point. I don't know much about getting this info out of AI, but I've seen atleast one video talking about team doing this successfully. It could basically give a heat map about what the AI was focusing on in a video game, what it saw as negative and what positive (how it would affect it's reward function)
That is to say, there is precedent in being able to get info out of AI's black box.
Video in question is Robert Miles "We were right! Real Inner Missaligment", and article later in the video about regognizing what AI thinks is "Understanding RL vision" by Distill (found in videos description or first result in google)
@@Trombi01 Yes. Makes sense that the 90% correct results could highlight the similar patterns or sequences that led to the result. These could then be taken up as avenues of research whereby organic chemists may one day be able to uncover causative principles behind the observed phenomena. Well, at least it’s a possibility.
But I wonder if maybe we’re rather like the economist in the joke where three men - an engineer, a chemist, and an economist - are stranded on a desert island with only a can of food. As dinner approaches they get hungry and begin discussing how to open the can. The chemist proposes that using a combination of salt water and other substances they may be able get it open through some corrosive process. The engineer suggests dropping the can from a platform (to be designed and constructed) of a given height. Finally, the economist begins his proposed genius solution with the words “let’s assume we have a can opener. Then …” Is the AI route like the economist’s solution, only we actually do already have the can opener? We’ll be able to open many cans, but not without the “already provided” can opener, i.e., the existing complex protein structures, and never actually figuring out origin and inner workings that led to the can opener in the first place.
Protein folding, that brings memories from the early 90 when I was working as a computer tech assistant for post docs who were working on this topic. Those were the days of having a 40k euro Silicon graphics workstation on your desk 😂
Hello Anton! So sorry for your loss hope you find some peace and appreciate you making these videos in difficult times. You never cease to amaze me 🙏
I have read papers invoking quantum mechanics to explain how proteins assemble so quickly given the complexity of this problem. They were difficult to understand but implied there is even more going on here than just regular chemistry.
Chemistry in general is about quantum mechanics
@@jyjjy7 but feynman said that there is not a material reality, only mathmatics exists; i dont believe in that serial sexual offender, i dont believe in his idealism, and i dont believe in quantum physics in general
No. That’s wrong, the field of computational biophysics does quite well with hardly a quantum model involved, and even more so with folding.
@@johnro8ot I don't know what you mean by that. The standard model of particle physics, literally the foundation of our understanding of the properties of particles, atoms and molecules, and how they behave and interact, is derived from quantum field theory.
@@jyjjy7 and what I mean is that Newtonian physics, called classical mechanics, accurately describes protein folding. There is a field of science called computational biophysics, look it up, and see what models people ACTUALLY use to simulate protein folding, and stop living in science la la land if you want to actually learn something.
Love the way you used "psilocybin" in the background when talking about medicine ("medical fields"). I believe it has the potential to be a big help to many people suffering from depression and possibly other mental illnesses.
And to connect to nature and the astral plane 😉
@Skippy dude 10mg of thc sends me to the psychward you got to use CBG to become more sensitive again.
It's nice that your Mom wanted you to be a doctor, but we have to be who we want to be. I'm sure she would be proud of you.
Most of us don’t have the luxury of being who we want to be. We have to cope with whatever arises.
@@SolveEtCoagula93 with a negative attitude, you never will.
@@Vicki_Benji 😂 😂
@@SolveEtCoagula93 only the lazy laugh. Again, your negative attitude is the problem.
@@Vicki_Benji 'Only the lazy laugh'? What a stupid statement. Where on Earth did you dig that up from? Any more deep insights into human nature - can't wait! 🙄
Maybe photosynthesis is the exceptional odd man out and chemosynthesis is the norm. This could mean that we simply don't know what to look for in way of a life signature... hence the word "alien".
Or, for some reasons or others, there are standard processes not seen in our neighborhood.
Or even...synthesis-synthesis. That might sound pretty deep, because it is, but think about it.
A principle one must consider is that if it can happen, it will. Where there's available energy the potential for some kind of life exists. For all we know, stars are conscious beings.
But even here on earth different types of photosynthesis have evolved.
This makes it more likely that it is not so uncommon, if life evolved somewhere else.
@@wolfgangkranek376 If there's 1 to the power of 300 different configurations then it seems there could certainly be more than just that one particular form which could do the same thing. Maybe what we have just happens to be the configuration that worked here 3.8 billion years ago.
You were a great son, remember that. I'll take you on as a second son, perhaps I could even be your grandmother at 77. A Doctor of science, emeritus, at Wake Forest University turned on my love for science, from the micro to the macro--I was getting my Masters in world religions when I went to one of his lectures and decided to take his class on The Universe as an elective. It was mind bending and quite glorious. That was in 1995 when I was 50; I have never since lost that joy of discovery. You have been nourishing me as of late. I keep up with everything in Science and Religion. Continue on
Doesn't the fact that the proteins fold autonomously, into amazing structures, make it more likely that life would be on other planets? Why would protein fold in amazing ways on our planet only?
I wonder if there are some principles of nature that make things like inevitable that we're simply not seeing yet
Answer to your question is in a way in your question, just that the premise in your question is a bit wrong I would say. You ask, "Why would protein fold in amazing ways on our planet only?". Well, this is our planet only because proteins happened to fold here in a certain way. If it would've happened in some other planet, it would be our planet and not this. And if proteins would've not folded in a certain way we wouldn't be here observing it to fold in a certain ways. I hope you get what I'm trying to say here? :)
@@NicholayN There are probably some mechanisms that prohibit many combinations from ever happening. Meaning that the combinations are wayyyy lesss than 10^300 or whatever. Also, maybe, some combinations of all these supposedly 10^300 can give legit functioning proteins with similar properties. There is not way scientists tested 10^300 combinations and only ONE was funtioning. You know what i mean. I see such narrow thinking by the scientists, but what do i know? i am not studying this field.
I've watched this 37 times and I can confirm this is a certified anton classic
I'm impressed by your time travelling abilities.
This episode feels a bit different, I think Anton has some deep beliefs about this topic, very interesting episode.
Deep beliefs?
@@Pop-zb3wr Exactly. To make an unprovable claim as fact shows very strong beliefs.
Yea idk why but it seems like he wants this to be true for some reason. Even though this “paradox” isn’t a fact at all
The resolution to the Levinthal paradox lies in the fact that protein folding is not a random process. Instead, it is guided by the amino acid sequence of the protein and the biophysical properties of amino acids, which drive the protein to fold along specific pathways. These pathways lead the protein to its native, functional structure more quickly than a random search would allow. This concept has been further supported by the development of the energy landscape theory and experimental observations.
I really think the moon and tides was instrumental. Because this means that any proteins and molecules get left high and dry for hours drying out, then wetted, then dried out, then wetted.
Oparin and Miller had the right idea, we just need to realize how rare a planet like ours with a GIANT moon, iron core so super duper radiation protection.
With that premise, planets with multiple moons and suns and faster tidal cycles would potentially arrive at viable biochemicals faster than our single sun/moon planet system.
What makes you think having a large moon is super rare? In out solar system it is but we dont know at all about giant moons of exoplanets.
@@28th_St_Air no. Multiple moons either don't have any influence because smaller and if larger would give too chaotic a total pattern without rhythm
@Fummy Because a moon as large as ours has actually never been found that I am aware of. No planet in our system had anything like the mass ratios earth/ moon does
What if there are other shapes that could lead to life in entirely different ways? Should we really chase after one specific one?
That's what crossed my mind also. And can life forms only exist if based on proteins? Maybe that requires a whole new theory that we can not even theorize on.
Yes, because if alternatives where possible we would have see them on this planet by now
@@Tubemanjac Sure, probably they are made out of magic
@@rommdan2716 did you really think through your argument? it's kind of bad
And why didn't we find something like that in our planet that's practically perfect?
i think this shows that there’s a possibility that life as we know or theorize may only be the tip of the iceberg.
I'm really glad I came across your channel. It's funny enough things that bother me, and when im searching for it and answers, it is rarely understood and the past couple of days your channel came accross that Really elaborate on the things i want to know about.. thanks for the content. You are awesome
Also important is, how close to that exact configuration does the sequence need to be to produce a functional RUBISCO analog. Maybe the gene only assumes the sequence it does because that is the most efficient form, but other alleles may be perfectly functional, just slightly less efficient.
Exactly! Of course the chances of nature folding "the one" Rubisco we know on Earth are insanely low, but who knows how many just as funtional, if not better analogs there are? In proteins, amino acids can be swapped out to some degree and it won't make them unfunctional. I assume that's the case not only with amino acid sequence, but folding as well.
If we truly are alone in the universe, there seems to be an awful waste of space... Plenty of room for humans to trim down our zero sum game behaviour and get out there.
Anton, I am glad that you are finding your way and while on your journey you chose to enrich our lives with every post.
Very inspiring and humbling.
Thank you.
we are not even in zero-sum game, but we behave like we are in one because we built a zero sum game economics systzm
Universe (and nature) never bets on one thing!
It's still statistically very unlikely that we are alone, none the less. But it may be that the conditions that allow for advanced lifeforms to develop are pretty rare, so maybe other civilizations existed or will exist but we'll never meet them because they are long gone or they still have to develop. It's even possibile that there are civiilizations in the same time frame as us, but so far that we can't really perceive them nor communicate.
actually, space is so big, we can't' even travel in a human life time to the nearest star. Sending humans to Mars and have them survive the trip and doing whatever we want on Mars with very high probability is already a huge task we don't know how to solve yet.
Satan said similar thing.
Only we are the waste of space.
Just as God gave us dominion over earth,he has given us the entire universe,we are still stuck here in this infernal prison.
So we made a machine to understand a process, and now we need to understand how the machine understands that process
I'm so glad you turned up in my program. It's been a while. Most count of how many T-shirts I bought when you were struggling.
Just an idea here. We do have a gazillion proteins, but we are still mostly Water. I love water, but I'm very picky about proteins. Plants only, despite something I learned from one of your videos! Keep shining!
Has there ever been an attempt to explain why bacteria grows so rapidly in outer space? Could it have anything to do with a certain kind of radiation?
This has always intrigued me and makes me wonder about the concept of panspermia. Interestingly enough, there's even an ancient Egyptian myth about life arriving to earth via asteroids.
"Has there ever been an attempt to explain why bacteria grows so rapidly in outer space?" Source?
@@harmless6813 It's a pretty well known fact that has been observed since space travel began.
@@deltasyn7434 Not really. I just checked and found tis paper:
"Effects of spaceflight and simulated microgravity on microbial growth and secondary metabolism"
Read it.
@@harmless6813 Sounds like you're trying to be a contrarian. This study confirms that the observation is correct as microgravity is what you have in outer space.
I took a wild guess about it having to do with radiation, but I never presented that as fact because I don't know.
That said, it seems that lack of gravity and cell motility is the determining factor causing the rapid spread of bacteria in space-like conditions. Which answers my question. Thank you.
@@deltasyn7434 As far as I can tell, there are _some_ bacteria that grow faster in space. Your original made it sound like bacteria do generally grow faster in space, which does not seem to be the case. Sorry if I misunderstood.
I think one solution to the paradox is, that the assumption of a free space around the protein, which would allow for the scanning of all possible configurations, is not true. The ribosome already restricts the free space a lot. Also the precursor amino acids are already having a certain conformation. They do not check out any other conformation, because they are already in their ground state structure. It seems to me similar to diffusion process which dimension is restricted, thus going on much faster than expected for the full dimensional space.
It was always recognised that life is a difficult thing to explain , just because you now have numbers to put it into a framework does not change anything . I love your videos , keep putting them out please.
Thanks! Anton! Your videos are amazing! And some how just listening to you soothes me! Instead of thinking about worldly issues! I focus my energy on gaining knowledge! Your so appreciated ❤
Abiogenisis itself may be a filter to life since we cant actually explain how it began. It's a very interesting topic given the environments the Earth has been through to even allow life to begin at all.
My guess is that the universe is carpet-bombed with simple life. Complex life springs up occasionally, and "intelligent" life is rare.
Human-level intelligence might be rare, but persistent complex life (I'm going off the genetic structuring aspect of eucaryotes here) probably leads to "intelligent" life proportionately common. Even a cricket is semi-intelligent, and those are so stupid that they'll eat each other while a tarantula is literally eating them both. The bigger aspect is more likely the amount of possible life, as Earth seems to be on the leading edge of a high-chemical-variety period of the universe: we are among the "early birds" in the universe.
@@absalomdraconis I might speculate the same. And there are man factors, such as the galactic goldilocks zone...our over all soup is richer than previous stellar generations. Cool Worlds speculates on this in fact.
It's only a guess, but I've always thought that life is common. Complex life, probably rarer. Intelligent life, quite rare indeed. I'm not entirely sure that humans count as "intelligent life," but I guess the next three hundred years or so should answer that question.
@@BronzeDragon133 My man we have technological civilization I think we are quite intelligent...
@@hubertino855 depends on how u see on it. Rats dsnt make "rat traps" but we make guns nukes etc
I have been working on this for many years, in the past, and there were no software tools able to provide any meaningful answer (I am a medicinal chemist, in any case). It seems now we have something working quite well in predicting protein structure. Explaining why, however, is another story. In any case, you don't necessarily need a specific protein to get a specific function. For instance, mammals use hemoglobin to carry oxygen; but some cephalopods use hemocyanine (based on copper instead of iron) to do the same. So I think to focus on a given protein, however relevant it may be to us, is not a good answer. Besides, we use 20 "natural" (to us) aminoacids, but somewhere else they could be more and different. Or may be no protein based life. We are still too much anthropocentric
Mr. Anton,
I enjoy listening to your dissertations and there are times when you define the probabilities in your subject in such a way that demands deep thoughts and this time is no exception.
Listening to this topic, what you've stated makes me want to apply what seems could be obvious, there was an intervention of sorts?
Just an old man pondering the evidence at the scene of of a profound event...
Have a good day.
It has been a long time since i studied this sort of thing, but i remember hydrophobic and hydrophilic interactions were great influences in the shapes of proteins, and these are in turn influenced by the properties and sequence of the individual amino acids. these interactions would greatly reduce the degrees of freedom. also the amino acid chains are built up one amino acid at a time so they would start folding as they went along. the rybosomes that assemble the amino acid chains into proteins are incredible "machines", they probably have (its been a while i cant remember for sure!) functions that also facilitate and guide the folding process. also there is certainly a whole host of additional proteins and sugars and nucleoties providing suport in the background. The AI is probably seeing patterns from the effects this additional cellular machinery has on folding, it will probably get better still.
I am wondering two things. First, are we assuming that of a given sequence, there is only one way it can fold for it to be functional? Of the 10^300 different combinations, could many of them be sustainable for life in a given circumstance? Second, is 10^300 the odds for that protein to fold a certain way at the given conditions here on Earth?
I think that number isn't that important. I'd guess the factors that influence WHY it folds is far more important, so the original number is relatively arbitrary.
I bet your mom would be very proud of you and what you have done with your ability to spread knowledge and empower people more than majority of doctors do.
Thank you for being you
Much love, mate. Thank you as always wonderful person, and YOU take care
The other factor contributing to the rare earth hypothesis is time. That is, even if the massively improbable circumstances happen twice, they may happen at completely different epochs.
Or different universes, or strata of universes within universes etc
Maybe other life has other forms of life out there have vastly different base configurations from our own, but if it's equally complex, then it doesn't really answer the question.
Exactly. Everyone says we've not found life anywhere else, but that may or may not be the case. We can only look for life that conforms to the only type of life we know - what exists on this planet.
Those proteins are built one amino acid at a time and don't just pop into existence. I guess their shape is pretty much determined by that sequence. Though this points to a precurser, maybe the RNA world and earlier chemical cycles that led to what we have today
Yeah, that part about deciding in milliseconds the right structure doesn't sound right, apart from chaperones. The sequence will limit the number of configurations possible very strictly and evolution will find a way to fine tune the structure or adapt it to other functions. Abiogenesis remains a mystery, but once you got something that reproduces, even if it's just a simple chemical reaction in a specific environment, it can take off from there.
That is interesting. Do proteins fold while they are assembling one amino acid at a time? If so, it seems that the next one in the sequence would be somewhat nonrandom and the next building block and folding event would happen at the same time following principles of least action or lowest energy or however you'd like to express it. Doesn't the shape at each depend mostly on geometry and whatever lowest energy state applies?
Can partly formed proteins somehow join to create bigger and more complex proteins? If this can happen, do they each retain whatever folding they had, or do they unfold and refold?
This wonderful person, is, well, wonderful.
Anton, you are one of the very few channels that I can learn something to talk about in conversions with friends and charm them with interesting knowledge.
I always though the most important part of the Drake equation was the "fraction of life-supporting planets that develop life". But we only had proof of one planet that could do that, Earth. Until we figure out how life can start, we will never know if it can start elsewhere.
Now that you attach a paradox to it, it makes it a lot more credible. In the next party I will say: "have you heard of the Levinthal paradox. No?! Ok, let me tell you."
The way you describe it, it's a wonder protein folding is so fast and consistent even just here on Earth.
So many ways to fold a protein . . . and yet they somehow know to fold in the correct manner over and over again. Something must be governing the process that we do not yet understand. Something that allows enormous speed and accuracy where neither should exist at all. Perhaps some influences from the subatomic level that make certain types of folding inevitable and others not.
We do pretty much understand it, and it is way less esoterical than you presume:
1) Amino acids (aas) can be grouped into water attracted and water repelled, so by swimming in their natural surroundig of a cell (water based) the water attracted aas will be attracted to the outside and the water repelled aas will be "clumped" on the inside of the protein.
2) "Chaperone" proteins exist that act like a scaffold or mold to help other proteins fold.
3) Due to 1) and 2) and also other factors, the correct shape of the amino acid is always an energetically stable form. This means folding is less like random movement and more like a ratchet mechanism, single steps can move towards the correct shape, but not towards an incorrect shape since this would require energy and the laws of theromdynamics forbid this.
@@mateolaskowski1731 Thanks for the explanation. Non-expert here, as you may well tell. But it reminds me of the molecular mechanisms of Mad Cow disease I've read about where rogue proteins cause other proteins to fold incorrectly in a pathological manner.
Similar to the situation with growing crystals of new compounds for x-ray diffraction. Nearly impossible until achieved, then inexplicably around the world, all of a sudden much easier.
@@JoeSmith-cy9wj That sounds a little supernatural to me. I'm afraid I'm going to have to invoke the Sagan Standard on this: extraordinary claims require extraordinary evidence.
Or does it just get easier because improving lab techniques are copied?
Please cite credible scientific sources for your claim. Peer-reviewed if possible, And no flakie social media citations.
@@douglascutler1037 I'm not here to debate anything. For the record, I hate religion, don't believe in god, and do believe in aliens. The crystal thin is just a known fact. How much hearsay and how much recorded evidence I have no idea. But the story has been around and prominent in that field for years. I'm sure Wiki might have something on it. Scientists seem to agree it's odd. I don't know if any statistics or data has been collected.
Anton has just started to grow his 'Mirror-Universe-Spock Beard'. I look forward to the results.
Anton is a great example of the rare Earth hypothesis. Where else in the universe could we find another Anton? Nowhere! 😃
Well if the universe is infinite, which it likely is because space is flat, then there’s infinite Anton’s out there
@@lionelmessisburner7393 Woo hoo! 😃
@@lionelmessisburner7393 the universe has to be infinite.
Well but the universe per see because that is just the finite limitation that we are aware of. But beyond that. It's infinity simply because how could something be absolutely finite? That isn't possible.
@@sonkeschmidt2027 I often thought of that too. But it could be something where it loops back on itself. When u get to the end of one side you just find yourself on the complete opposite side. I don’t think that’s the case though, likely it’s infinite
@@lionelmessisburner7393 even if you loop it, it would be infinite. Because the looping would never end.
Except how do you know you are looping? In order to know that you must remember that you've been there before. Without that memory you wouldn't even know that you were looping, you would just go on infinitely.
Spent years driving into work listening to the Drill. Nice finally working from home so I can now see these guys live.
I'm extremely skeptical of the 300 zeros theory. It would mean that things are totally random and yet they aren't. As we all know, we now have AI that can predict quite well how something will fold. This can only mean that such folding is very deterministic, or else AI would output random garbage