Harmonize publicly available scRNA-seq datasets for the eye
The code base that underlies the data for plae.nei.nih.gov
- scANVI (scVI version 0.13)
- 4000 HVG
- 5 epochs
- 15 latent dimensions
- config.yaml for SnakeQUANT and SnakePOP
- configSCEIAD.yaml for SnakeSCEIAD
- 44 datasets
- 4 species (human, mouse, macaque, chicken)
- 1,136,041 cells
- ~60 different cell types
- scEiaD database for PLAE
- Seurat (v3)
- anndata (h5ad)
- Cell Level Metadata
- Cell Level Metadata in FST format used by web app
- Counts (Rdata sparse matrix)
Visit plae.nei.nih.gov -> Data for differential gene expression tables, study level counts, and more.
It is possible! Even better, you don't have to ask me! Or tell me!