Thread by @sinabooeshaghi

1/ Worried to present @GoogleColab Python tutorials & R vignettes for complete #scRNAseq. Thanks to @satijalab @fabian_theis, @pmelsted, @lpachter @LambdaMoses @lioscro et al. and
biorxiv.org/content/10.110…
nature.com/articles/nbt.4…
genomebiology.biomedcentral.com/articles/10.11…
See
kallistobus.tools/tutorials

@GoogleColab @satijalab @fabian_theis @pmelsted @lpachter @LambdaMoses @lioscro 2/ We've tried hard to make sure that the Python tutorials == R vignettes, so that language choice does not affect method. But we have not been able to get exact replicability.

@GoogleColab @satijalab @fabian_theis @pmelsted @lpachter @LambdaMoses @lioscro 3/ For example, we've learned that Scanpy flavor="cell_ranger" is consistent with Seurat but flavor=“seurat” is not consistent with Seurat for highly variable gene detection.

See example python notebook: colab.research.google.com/github/pachter…

and example R notebook: colab.research.google.com/github/pachter…

@GoogleColab @satijalab @fabian_theis @pmelsted @lpachter @LambdaMoses @lioscro 4/ Even the order in which steps are performed can work with one software platform, and not the other. In addition we have had a lot of trouble getting a scater analysis (@davisjmcc et al.) to agree with Scanpy.

@GoogleColab @satijalab @fabian_theis @pmelsted @lpachter @LambdaMoses @lioscro @davisjmcc 5/ We did find that @GoogleColab is very useful: you can verify for yourself that things work using only a web browser and it provides full transparency. If you're frustrated with tying to install Seurat after its removal from CRAN you'll learn how to do so directly from @GitHub.

@GoogleColab @satijalab @fabian_theis @pmelsted @lpachter @LambdaMoses @lioscro @davisjmcc @github 6/6 We'll be happy to accept feedback in any form, and new features / demonstrations will have to consist of pull requests to *both* the Python tutorials and R vignettes. github.com/pachterlab/kal…
Particularly interested in help to get the Python and R notebooks reconciled.

Excited to say that Seurat should be back on CRAN imminently, per @satijalab. We will update our notebooks once it is.