Hi Ana,
Maybe I'm misunderstanding but you probably should do QC on each
chromosome 1-22 (as you've done for chr6), thin them by LD, then join
into one file and do PCA on that.
FlashPCA2 can't read the new pgen/bpgen file format so you'll need to
convert the final dataset to the old bed/bim/fam format.
The exclusion_regions_hg19.txt file is based on hg19 positions, if
your data is in a different build then you'll need different
positions. It's intended for general purpose PCA, i.e., excluding high
LD regions that can mask population structure. But you can change it
if that's not what you want.
Regards
Gad
> --
> You received this message because you are subscribed to the Google Groups "flashpca-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
flashpca-user...@googlegroups.com.
> To view this discussion on the web, visit
https://groups.google.com/d/msgid/flashpca-users/5138403e-92f9-4a15-89ff-f62e16c59e98%40googlegroups.com.