Inflated multivariate GWAS pvalues

[Shannon Cusack]

Hi,

I'll echo other's sentiments and thank you all for the great package. 

I was successfully able to run a multivariate GWAS on the three factors from our user-specified model. However, the results do not look trustworthy and I am wondering if it is a user issue vs. something else that may be going on. Specifically, for two of our factors, thousands of SNPs are meeting the genome-wide significance level, when we would only expect a handful or so to meet. Again, I am guessing this is a user error that occurred somewhere on my end, but thought I would check in just in case.

Thank you for your time, 

Shannon

[Elliot Tucker-Drob]

It's hard to say what's going on without knowing specifics. One thing that we've seen is that this can happen when there is a perfect linear dependency between the measures used. For instance, including grey matter volume, white matter volume, and total brain volume (grey+white matter) may produce this.Of course that's one of many possibilities. It could be anything from an error in processinging/Qcing the sumstats, to uncontrolled popstrat, to using indicators with minimal h2 or near 0 genetic correlation, to model misspecification.

--
You received this message because you are subscribed to the Google Groups "Genomic SEM Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to genomic-sem-us...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/genomic-sem-users/c0df0858-7300-4e49-8bb1-b3e5077a8e41n%40googlegroups.com.

[Michel Nivard]

Hi Shannon, 

Could you share the model output (without SNPs) so we can try to the set of the potential causes of this behavior.

If your for any reason not able or willing to share the substantive results feel free to replace the trait names with random names.

Michel 

Op wo 7 feb. 2024 om 18:12 schreef Elliot Tucker-Drob <tucke...@gmail.com>

To view this discussion on the web visit https://groups.google.com/d/msgid/genomic-sem-users/CAH2Be0uabwGKqDBf6v5YHXR5KFRxH3Y6%3DSr6Lv5SXOvZccyabw%40mail.gmail.com.