search

cozygene / bisque

An R toolkit for estimation of cell composition from bulk expression data
68 stars 20 forks source link

Include covariates for residual CPM #11

Open quattro opened 4 years ago

quattro commented 4 years ago

Hi all, thanks for developing this wonderful tool--it is SO much faster than existing methods and very intuitive!

I have a feature request: is it possible to add optional covariate parameters to the primary inference functions? The idea is to use these features to residualize the CPMs in the bulk data, before matching with reference panel data (or within the reference-free approach).

Off hand, one approach would be to estimate logCPMs, regress out covariates, then take exp of the residuals to place values back into CPM-space.

thanks!

brandonjew commented 4 years ago

Hi Nick,

Thanks, glad to hear it!

Covariate adjustment is something a few people have expressed interest in. I'll implement this (or something similar, still need to discuss this with others) as a test feature in the next week or so and include it in the github version of the package. I'll send you an update here when it's updated.

Thanks again and let me know if you have any other questions or suggestions. Brandon