data_requirements.md

Notes on the types of data we need.

The default settings are tailored toward two pre-set states "Carcinoma" and "Lymphoma" but we could define our own custom ones if we liked.

Discovery / Recovery from bulk RNA-seq data

1. Expression Matrix - A TSV of format, with TPM or CPM counts (anything that is normalized really...)

   target	sample1	sample2	...
   geneA	123.1	123.8	...
   ...	...	...	...
   ...	...	...	...

2. Annotation File - A TSV of format

   sample	grouping	other data	...
   sample1	group1	...	...
   sample2	group1	...	...
   ...	...	...	...

Recovery from scRNA-seq data

1. Expression Matrix - A TSV of format

   target	sample1	sample2	...
   geneA	123.1	123.8	...
   ...	...	...	...
   ...	...	...	...

2. Annotation File - A TSV of format

   sample	grouping	other data	...
   sample1	group1	...	...
   sample2	group1	...	...
   ...	...	...	...
   However, the grouping here is more restricted.
   Since we are working with Carcinomas the "grouping" column is Celltype which supports the following subtypes:

   • 'B.cells'
   • 'CD4.T.cells'
   • 'CD8.T.cells',
   • 'Dendritic.cells'
   • 'Endothelial.cells',
   • 'Epithelial.cells'
   • 'Fibroblasts'
   • 'Mast.cells
   • 'Monocytes.and.Macrophages'
   • 'NK.cells'
   • 'PCs'
   • 'PMNs'

Discovery from scRNA-Seq data

Uses again the same two formatted TSVs alongside with a configuration yml file.