search

weizhouUMICH / SAIGE

GNU Lesser General Public License v3.0
188 stars 73 forks source link

Pan-UKB chrX data looks weird #252

Closed Shicheng-Guo closed 3 years ago

Shicheng-Guo commented 3 years ago

Dear Wei,

I notice you are also the key team member of Pan-UKB. I find a weird thing. I checked there were 4 million SNPs in chrX in UKB imputation data, however, only about 913,562 SNPs in summary statistics from Pan-UKB project, >75% are lost in the final summary statistics indicating there is very strong filtering for chrX. Really want to know what kind of criterion could make such strong selection in chrX. Hope to receive some reply. Thanks.

Shicheng

image

weizhou0 commented 3 years ago

Dear @Shicheng-Guo,

Thanks for the question! Will get back to you asap.

Wei