6. Dimensionality reduction of scRNA-seq data

This lecture by Paulo Czarnewski (NBIS, ELIXIR-SE) is part of the course "Single cell RNA-seq data analysis with R" (27.-29.5.2019). Please see www.csc.fi/web/training/-/scr... for the full course description and all the materials.
PCA : 5:31
tSNE : 17:05
UMAP : 32:42
Link shown in the tSNE slide to "How to Use t-SNE Effectively": distill.pub/2016/misread-tsne/

[@dokim2242]

PCA : 5:31
tSNE : 17:05
UMAP : 32:42

[@ChipsterTutorials]

Thank you! Added these to the video description as well!

[@pl1840]

Concerning the slide at 28:18, I would like to point out that the only stochastic element in tSNE is the initialisation of the low-dimensional embedding: even the accelerated BH algorithm is not stochastic. In other words, tSNE is just as stochastic as UMAP: the result will be the same given the same initialisation.