Meta-analysis of the glm results conditioned upon the top SNP

[Jingjing Bai]

Dear Chris,

I am trying to meta-analyze the outputs which were conditioned upon the top SNP using plink2.0.

I am wondering if plink1.9 will automatically pool the adjusted estimates with TEST column showing ADD or I need to extract the rows before meta-analysis. Because in the conditioned output, I found for one specific SNP, there are two rows with different statistcs. Their TEST column are different, one showing ADD and the other is the top SNP that I included. Just curious how plink handles these for meta-analysis.

Many thanks,

Jingjing

[Chris Chang]

From the plink 1.9 --meta-analysis documentation:

"- If a variant appears more than once in the same file (e.g. --linear/--logistic output), only the first appearance is considered. Add the --meta-analysis-report-dups flag if you want the later appearances to be logged in the .prob file and included in the 'problematic line' count."

--
You received this message because you are subscribed to the Google Groups "plink2-users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plink2-users...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/plink2-users/7281a2b4-ac0f-411c-abcd-b31da8f8e9c2n%40googlegroups.com.