Open mglev1n opened 1 year ago
Hi @mglev1n! There are some inaccuracies that we've experienced with that branch in our testing (compared to PRScsx), hence why it hasn't yet been merged.
There are some inaccuracies that we've experienced with that branch in our testing (compared to PRScsx)
Did you have a sense of where/why this might be occurring? Happy to take a stab at troubleshooting
No idea at all (it's probably somewhere in the MCMC code, which is large); we're not entirely familiar with the revised algorithm in PRScsx, so it's hard to figure out where things are going wrong. Any help is appreciated!
It looks like there is a branch
abf/transancestry
that provides the multiancestry PRS functionality included in PRScsx (https://github.com/getian107/PRScsx). In some simple testing this branch appears to function correctly, but was curious about the official status of this branch? Thanks for the great package/optimizations.