search

gersteinlab / PsychENCODE_SingleCell_Integrative

Integrative Analyses across 388 human brain samples at single-cell resolution
Other
10 stars 2 forks source link

Some highly expressed genes were filtered out from the pseudo-bulk normalized expression values #4

Closed patzaw closed 2 months ago

patzaw commented 2 months ago

Hi,

I recomputed pseudo-bulk counts per cell type starting from the counts you provided in the annotated matrices.

I found a few hundred genes that are well expressed but are not present in any of your pseudo-bulk matrices.

Here are the top 20 genes that I found highly expressed in the samples but are not present in your pseudo-bulk matrices: IL1RAPL1, FTX, DMD, LINC00632, DANT2, ATRX, TTTY14, AL008633.1, GPM6B, LINC01278, NLGN4Y, HDAC8, USP9Y, TSPAN7, IDS, CDKL5, UTY, JPX, CASK, DIAPH2.

Do you know why they have been filtered out? I believe they pass your filtering criteria described in the Methods and Materials, but I might have missed something.

Thank you,

Patrice

prashantemani commented 2 months ago

Hi Patrice, That is an interesting observation. I have reached out to my colleague who worked on those matrices, and will try to have a conversation with him before I get back to you. Best, Prashant

prashantemani commented 2 months ago

Hi Patrice, My colleague Declan looked into this carefully: All the genes you mentioned are on the X or Y chromosomes, which were not analyzed in the associated QTL study. That is why you were unable to find them. Hope this helps! Prashant

patzaw commented 2 months ago

Hi Prashant

Thank you for this information. It does help indeed. Maybe you could add this info to the description of the file for future users.

Thanks again

Patrice