morris-lab / CellOracle

This is the alpha version of the CellOracle package
Other
302 stars 53 forks source link

Does celloracle need pseudotime information to give cluster wise GRNs? #204

Open ruchikabhat opened 2 months ago

ruchikabhat commented 2 months ago

I have a scRNA seq data for mouse and got a pseudotime plot using this tutorial: https://github.com/mousepixels/sanbomics_scripts/blob/main/celloracle_pseudotime_GRN.ipynb But the resultant graph looks reversed as per the literature knowledge of precursor cells. The list of GRNs filtered based on one cluster +one condition together (as I provided the metadata with combined info for the get.linksstep) shows coef.mean values for certain target genes as negative whereas their expression on the Dotplot from scRNA data shows reverse trend. Forexample Nur44 celltype in my scRNA seq data coming from Adult mouse show low expression of Cxcl3 but the source target GRN list for AdultNur44 shows Nfkb1 is linked to Cxcl3 with +coef_mean value. How to interpret these results?