Severe p-value inflation in BLINK with 402 individuals and 1.8M SNPs — how to address?

[Ido Lybovits]

Hi everyone,

I’m running GWAS using BLINK C on a dataset with:

• 402 individuals

• ~1.8 million SNPs

• PCA covariates included to account for population structure

Despite this, I’m encountering severe p-value inflation—my QQ plots show a steep upward line, and I get thousands of seemingly significant SNPs that don’t appear biologically meaningful.

Any advice or insights would be greatly appreciated!
Thanks in advance,
Ido

Validation I made:
1. files format are correct
2. all NA changed to NaN
3. I working with --vcf v format