jianyangqt / gcta

GCTA software
GNU General Public License v3.0
87 stars 26 forks source link

Regarding the issue of default data filtering in FastGWA #97

Closed yuxizhang01 closed 2 months ago

yuxizhang01 commented 2 months ago

Hello, When performing GWAS analysis using FastGWA in GCTA, my input genotype data contains 60 million SNPs. However, after automatic filtering with the default parameters, only about 40 million SNPs remain. I tried to resolve this by relaxing the missing rate cutoff, but since the cutoff cannot be greater than or equal to 1, some SNPs are still being filtered out.

Are there any parameters that allow for no data filtering during GWAS, and instead perform pure association analysis, or are there other methods to retain all of my SNPs?

Attached is a screenshot of my standard output. 微信图片_20240906091811

longmanz commented 2 months ago

Hi, You can use flag --nofilter to mute any filtering steps (which is equivalent to --maf 0 --geno 0). Note that the --geno flag indicates the genotyping rate, so if you set it to --geno 0.99999 it will filter out all those variants with genotyping rate less than 0.99999. That is probably the reason why some variants are missing in your analysis.