Видео

I tried to run this script, only faced 1 issue with 'i' object not found. It has to be defined first before we include it in for loop.

Thank you for sharing. Where I can download the human_mouse_gene_orthologues_csv?

If you are using Seurat V5: at 3:14 instead of the code MergedNML@assays[["RNA"]]@data@x you can write MergedNML[["RNA"]]@layers[["counts.NML_I"]]@x This will show the matrix for NML_I, and you can view the others by changing NML_I to NML_II or NML_III. At 6:10 instead of MergedNML@assays[["RNA"]]@var.features you can write VariableFeatures(MergedNML)

Hello Professor, I am working on scRNA data by following your videos I get an error on PCA analysis. MergedNML <- RunPCA(MergedNML) Error in PrepDR(object = object, features = features, verbose = verbose) : Variable features haven't been set. Run FindVariableFeatures() or provide a vector of feature names. How to resolve this error. I am using Seurat version 4

Great video! Out of curiosity do you know if there is a way to perform CellRank using integrated data?

I am getting error while doing adjacency matrix. Cannot allocate a vector of size 37gb how to solve it? I am working with 16gb ram/ 1TB storage laptop

Thank you! This helps me a lot! I successfully changed the name but when I do a subsequent analysis in my seurat, I get an error message that says, subscript out of bounds. Is this caused by genes without equivalent annotation? How to troubleshoot this?

quick question: shouldn't GSEA be applied on the complete set of genes with respective logFC? here, FindMarkers function produces a matrix with only differentiated genes across conditions.

it is very helpful

I think in minute 23 you got the same cluster becosue you are graphin using the idents from seurat not the clusters generated by monocle.

You are a true hero professor! <3

Thanks ❤

hi, can i ask question i encountered problem which some cells annotated as epithelial cells in level 1 annnotation by azimuth but if i look upon the finest level annotation, some of those epithelial cells (in level 1), annotated as non epithelial cells (like cd8, cd4, endothelials, and etc). though the amount of cell population that may misidientified is significantly few compared to majority of the cells when i look upon cannonical markers, it seems like some of the cells belong to epithelial cells are not expressing EPCAM how to solve this problem?

How do you convert csv file into seurat object?

How do you convert loaded csv files in R to seurat format? In this case do you still need the Read10x() function? Do I put the csv into seurat object somehow by CreateSeuratObject?

Thank you for the tutorial! How can I download scripts from akoyabio?

hello, it is really help me for my analysis. can I request that you can make a video which comparing the gene expression between two different species, like mouse and human. I wanna compare my subset fibroblast cells between mouse and human. But I don't know how. thank you~

Professor, thanks for your video. I have one question, could i use scDblfinder with mergerd data? Thanks.

Thank you for this clear and easy-to-follow tutorial! I have a query: I have an RDS file that contains a list, where one part is an RNA Seurat object and the other is a dataframe. I need to convert this RDS file into an HDF5 format for downstream analysis. Would it be okay to follow this tutorial to convert just the Seurat object to an h5ad file, and handle the dataframe separately? Or is there a recommended way to include both components in the conversion process?

I do 4 groups ,

Thanks professor. if you could share the notebook, that would be very helpful. Thank you for your effort !

How come your computational speed is so fast! Care to share configurations?

Hi, thank you so much for the videos. It helps me a lottttt!!!!🤩🤩

How can I modify the seurat object's barcode to be the same as combined TCR-seq's barcode?

Thank you so much Dr!

Hi , Could you please do a video to teach Bioconductor SummarisedExperiment S4

Looking forward to more analysis using super computer/linux for sequencing

thank you professor the video is so nice, i wonder if you can add the subtitle for the vid. This vid help me so much

Love it ❤. You are the light of beginners. Could you do more on Super Computer.

Probably it is a bit off-topic here. If you have time, could you have some tutorials on how to use R and python in Super Computer? I have access to Super computer at workplace, but the HPC team didn't provide support on how to use R and Python in super computer to analyse data and sequencing data. You may have good tips and would be helpful to viewers.

If i downloaded dataa from geo dataset with barcode.tsv, matrix.mtx, fetaure.tsv and image.jpg.. how to create seurat object and process

What a saviour

I am having a doubt.. i have 6 sample and i done all procedure including intergration and find clusters etc.. i got about 26 clusters..further I proceeded with find marker genes for each clusters ..is that means 26 cluster will be 26 different cell type.. can i run this type of feature plots based on marker genes apart from find marker function and can chnage the cluster name....pls help to rectify the same..

How to split them and rename into reads and technical runs?

Thank you for the helpful video! I got weird result when do cell type annotation using seurat method. May I know if we can apply cell2location output to cellchat? Thank you so much!

library(STutility) # built to work on top of Seurat, NMF implemented as RunNMF library(tidyverse) library(RColorBrewer) ### 1. Prepare data: 10X Visium data samples <- "data/V1_Adult_Mouse_Brain_filtered_feature_bc_matrix.h5" spotfiles <- "data/tissue_positions.csv" json <- "data/scalefactors_json.json" imgs <- "data/tissue_hires_image.png" infoTable <- data.frame(samples, imgs, spotfiles, json) se <- InputFromTable(infoTable, min.gene.count = 100, min.gene.spots = 5, min.spot.count = 500, platform = "Visium") cscale <- c("lightgray", "mistyrose", "red", "darkred", "black") ST.FeaturePlot(se, features = c("nFeature_RNA","nCount_RNA"), cols = cscale, ncol = 2, pt.size = 1.85) ### 2. Load the image se <- LoadImages(se, time.resolve = FALSE, verbose = TRUE) ImagePlot(se, method = "raster", type = "raw") se <- MaskImages(object = se) ImagePlot(se, method = "raster", type = "masked") # Masked image ### 3. NormalizeData, FindVariableFeatures and ScaleData se <- SCTransform(se) view(se@meta.data) ST.FeaturePlot(se, features = c("Cck", "Mbp"), pt.size = 2, cols = cscale) FeatureOverlay(se, features = c("Cck", "Mbp"), cols = cscale, pt.size = 2, type = "raw") ### 4. non-negative matrix factorization (NMF) dimensional reduction se <- RunNMF(se, nfactors = 40) # choose the number of factors, default = 20 ST.DimPlot(se, dims = 4:5, # 1:40 ncol = 2, grid.ncol = 2, reduction = "NMF", pt.size = 1.5, center.zero = F, cols = cscale, show.sb = FALSE) # print a summary of the genes that contribute most to the dimensionality reduction vectors print(se[["NMF"]]) # barplot to summarize the top most contributing genes for each factor FactorGeneLoadingPlot(se, factor = 1) ### 5. Clustering se <- FindNeighbors(object = se, verbose = T, reduction = "NMF", dims = 1:40) se <- FindClusters(object = se, reduction = "NMF", verbose = T) view(se@meta.data) VlnPlot(se, features = c("Olig1", "Th"), group.by = "seurat_clusters") # plot the clusters spatially qual_col_pals = brewer.pal.info[brewer.pal.info$category == 'qual',] col_vector = unlist(mapply(brewer.pal, qual_col_pals$maxcolors, rownames(qual_col_pals))) ST.FeaturePlot(object = se, features = "seurat_clusters", cols = col_vector, pt.size = 2)

Could you please help me I am following your tutorial and when I get to the count.sums, I cannot go any further. In the table there seems to be a normalizing of the cell numbers. I have over 10000 cells in some clusters but when I look at the number of cells present in the table I get 100, 500 etc. Somehow the cell number is not translating the proper raw cell numbers so when I run the dot plot I do not get proper sizing

good course!!!

Thank you so much for the helpful videos. I watched your channel for a long time. When I run FindTransferAnchors(), I got Warning: Layer counts isn't present in the assay object; returning NULL. And many cell types suppose to present when using SpatialFeaturePlot() have zero score. Can't upload a plot here. Would you please have a suggestion? Thank you so much!

Thank you so much for posting this. Hope to see more of cellranger

could not read H5, need to install hdf5r, how to solve this problems

Hello and respect Special thanks for your useful videos I have an important question. I would appreciate it if you could help me I have performed spatial transcriptome data analysis (Visium HD) for colorectal cancer as instructed in your videos If we analyze the spatial transcriptomics data and want to continue the QuPath analysis, how should we find the file related to the QuPath software that is related to colorectal cancer? And related to the results of our analysis? I mean, how can we find the qptiff file that you upload in QuPath software for colorectal cancer? With great respect

I dont have Library(tidyverse), Any suggestion?

Is this a pipeline for Xenium data alone, or includes the single-cell RNA data in it ?

I have one more doubt, if we are using only one sample, then can we use fgsea, is it make any sense.

How can I analyse single .tsv file on R using Seurat? e.g. GEO - GSE249493

library(Seurat) library(spacexr) library(tidyverse) set.seed(12345) ### Load Visium HD data cortex <- Load10X_Spatial(data.dir = "../Seurat/Visium_HD/", bin.size = 8) # bin.size = C(2,8,16) cortex <- NormalizeData(cortex) cortex <- FindVariableFeatures(cortex) cortex <- ScaleData(cortex) # sketch the cortical subset of the Visium HD dataset cortex <- SketchData(object = cortex, ncells = 50000, method = "LeverageScore", sketched.assay = "sketch") DefaultAssay(cortex) <- "sketch" cortex <- FindVariableFeatures(cortex) cortex <- ScaleData(cortex) cortex <- RunPCA(cortex, assay = "sketch", reduction.name = "pca.cortex.sketch", verbose = T) cortex <- FindNeighbors(cortex, reduction = "pca.cortex.sketch", dims = 1:50) cortex <- FindClusters(cortex, cluster.name = "seurat_cluster.sketched") cortex <- RunUMAP(cortex, reduction = "pca.cortex.sketch", reduction.name = "umap.cortex.sketch", return.model = T, dims = 1:50, verbose = T) DimPlot(cortex, label = T) # create the RCTD query object using 'SpatialRNA' function counts_hd <- cortex[["sketch"]]$counts cortex_cells_hd <- colnames(cortex[["sketch"]]) coords <- GetTissueCoordinates(cortex)[cortex_cells_hd, 1:2] query <- SpatialRNA(coords, counts_hd, colSums(counts_hd)) ### load in a scRNA-seq reference dataset ref <- readRDS("../Seurat/Visium_HD/allen_scRNAseq_ref.Rds") counts <- ref[["RNA"]]$counts view(ref@meta.data) cluster <- as.factor(ref$subclass_label) levels(cluster) <- gsub("/", "-", levels(cluster)) # cluster <- droplevels(cluster) nUMI <- ref$nCount_RNA # create the RCTD reference object reference <- Reference(counts, cluster, nUMI) # Creating RCTD Object RCTD <- create.RCTD(query, reference, max_cores = 8) # max_cores: for parallel processing. # run RCTD RCTD <- run.RCTD(RCTD, doublet_mode = "doublet") # add results back to Seurat object cortex <- AddMetaData(cortex, metadata = RCTD@results$results_df) view(cortex@meta.data) DimPlot(cortex, label = T) DimPlot(cortex, group.by = "first_type", label = T) # change NA to Unknown cortex$first_type <- as.character(cortex$first_type) cortex$first_type[is.na(cortex$first_type)] <- "Unknown" DimPlot(cortex, group.by = "first_type", label = T) # project RCTD labels from sketched cortical cells to all cortical cells cortex <- ProjectData(object = cortex, assay = "Spatial.008um", full.reduction = "pca.cortex", sketched.assay = "sketch", sketched.reduction = "pca.cortex.sketch", umap.model = "umap.cortex.sketch", dims = 1:50, refdata = list(full_first_type = "first_type")) DefaultAssay(cortex) <- "Spatial.008um" view(cortex@meta.data) DimPlot(cortex, label = T, group.by = "full_first_type") SpatialDimPlot(cortex, group.by = "full_first_type", label = T, repel = T, label.size = 4)

professor, can u help me with this Integrating data Warning: Layer counts isn't present in the assay object; returning NULL Merging dataset 5 4 into 6 Extracting anchors for merged samples Finding integration vectors Error: std::bad_alloc

Very useful work sir. Continue doing it.

Professor,how do we get only 2 plot- control and IPF,I'm getting 6 of them