GWAS to MTAG adjustment factor

[gina-parcesepe]

I've followed previous advice of running the single trait MTAG to work out the factor that the MTAG beta's and SE's need to be multiplied by to compare to the GWAS results.
The mean factor (MTAG_beta/GWAS_beta) is looking fine but for one of my traits the range for the factor is [0.03, 70.08].
The maxFDR for this trait is ~20% (but was ~19% in the original GWAS).

I'm not sure whether I've understood what you've said on previous issues on this matter correctly?

[paturley]

Hi. I'm confused by why there is a range. Isn't the single-trait MTAG roughly a constant multiple of the original GWAS summary statistics across all SNPs?

…

On Tue, Jun 11, 2024 at 8:40 AM gina-parcesepe ***@***.***> wrote: I've followed previous advice of running the single trait MTAG to work out the factor that the MTAG beta's and SE's need to be multiplied by to compare to the GWAS results. The mean factor (MTAG_beta/GWAS_beta) is looking fine but for one of my traits the range for the factor is [0.03, 70.08]. The maxFDR for this trait is ~20% (but was ~19% in the original GWAS). I'm not sure whether I've understood what you've said on previous issues on this matter correctly? — Reply to this email directly, view it on GitHub <#214>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AFBUB5POGBPHBFAJUYWR4I3ZG3V3VAVCNFSM6AAAAABJEIERGSVHI2DSMVQWIX3LMV43ASLTON2WKOZSGM2DMMZSHA2TKOI> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[gina-parcesepe]

Hi, it is not roughly the same for one of my traits. Calculated by trait1$mtag_beta / trait2$gwas_beta , I had the following for mean [range]:

trait 1: 0.11 [0.08, 0.13]
trait 2: 0.15 [0.03, 70.08]
trait 3: 0.18 [0.04, 0.68]