search

Katsevich-Lab / sceptre

An R package for single-cell CRISPR screen data analysis emphasizing statistical rigor, massive scalability, and ease of use.
https://katsevich-lab.github.io/sceptre/
GNU General Public License v3.0
26 stars 8 forks source link

Question about full output #34

Closed nchernia closed 1 year ago

nchernia commented 1 year ago

Thank you for the great tool! It is really nice to have a tool well-grounded in statistics that is also easy to use and reasonably fast.

I wanted to see the coefficients of the GLM and so ran run_sceptre_highmoi with full_output = TRUE, and I see promising fields such as xi, omega, alpha, nu. What do these correspond to? Is there a way to get at the question of which confounders are most important?

ekatsevi commented 1 year ago

Thanks for your question. Unfortunately, SCEPTRE does not currently give insight into which confounders are the most important. This would be an interesting functionality to add to a future iteration of our software. The fields xi, omega, alpha, nu refer to the parameters of the fitted skew-t distribution; see the documentation at ?sn::dst for more information on these.

Also, thanks for your comment that SCEPTRE is reasonably fast! I wanted to let you know that, at present, SCEPTRE (low-MOI) is orders of magnitude faster than SCEPTRE (high-MOI). We intend to upgrade the latter functionality to match the speed of the former in the coming weeks.