search

martinjzhang / scDRS

Single-cell disease relevance score (scDRS)
https://martinjzhang.github.io/scDRS/
MIT License
110 stars 15 forks source link

Issues with generating gs files #38

Closed Li-mengjie closed 1 year ago

Li-mengjie commented 2 years ago

Hello, when I used this software to generate xx.gs files with weights, I found that the xx.gs file contained many negative values. Is this normal ?Such as: ZSCAN31:-0.16077,CMC2:-0.17637,PRH1:-0.17909,FDXR:-0.18069,PROM2:-0.18672,IFI35:-0.19512,AP001029.1:-0.19552,AC087500.2:-0.19587,LINC00642:-0.20189,TLCD3A:-0.20189,AC011468.5:-0.21043. image

martinjzhang commented 2 years ago

Hi,

This is not normal. The weights should be all positive. I couldn't think of a way that the software can output negative weights. Could you send us a minimal reproducible example?

scDRS uses one-sided z-scores where negative values correspond to non-significant p-values (>0.5). If your z-scores are two-sided, you can convert them to p-values and input the p-value file to scdrs munge-gs.