Fewer significant sites after MTAG analysis

[KKFFCC]

When I select traits with higher genetic correlation for MTAG analysis, there are a lot less significant sites in each trait in the output file. What is the reason?

[paturley]

Hello,

There could be many reasons for this. It is very difficult to answer this question without seeing the log files corresponding to your analyses.

Best, Patrick

On Tue, Nov 30, 2021 at 8:37 PM KKFFCC @.***> wrote:

When I select traits with higher genetic correlation for MTAG analysis, there are a lot less significant sites in each trait in the output file. What is the reason?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/JonJala/mtag/issues/148, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFBUB5JJUI227L3N7ZFNHUDUOV355ANCNFSM5JDH5UNQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[KKFFCC]

Hi,Patrick This is the .log file I got from running. ###############log##################### Dropped 1247011 SNPs due to strand ambiguity, 6787888 SNPs remain in intersection after merging trait1 Dropped 1072 SNPs due to inconsistent allele pairs from phenotype 2. 6016301 SNPs remain. Flipped the signs of of 9716 SNPs to make them consistent with the effect allele orderings of the first trait. Dropped 0 SNPs due to strand ambiguity, 6016301 SNPs remain in intersection after merging trait2 Dropped 578 SNPs due to inconsistent allele pairs from phenotype 3. 5826636 SNPs remain. Flipped the signs of of 2575794 SNPs to make them consistent with the effect allele orderings of the first trait. Dropped 0 SNPs due to strand ambiguity, 5826636 SNPs remain in intersection after merging trait3 ... Merge of GWAS summary statistics complete. Number of SNPs: 5826636 Using 5826636 SNPs to estimate Omega (0 SNPs excluded due to strand ambiguity) Estimating sigma.. Checking for positive definiteness .. Sigma hat: [[1.014 0.068 0.016] [0.068 1.099 0.068] [0.016 0.068 0.918]] Mean chi^2 of SNPs used to estimate Omega is low for some SNPsMTAG may not perform well in this situation. Beginning estimation of Omega ... Using GMM estimator of Omega .. Checking for positive definiteness .. Completed estimation of Omega ... Beginning MTAG calculations... ... Completed MTAG calculations. Writing Phenotype 1 to file ... Writing Phenotype 2 to file ... Writing Phenotype 3 to file ... Summary of MTAG results:

Trait # SNPs used N (max) N (mean) GWAS mean chi^2 MTAG mean chi^2 GWAS equiv. (max) N 1 ...soc.fastGWA_rsid 5826636 15349 15259 1.085 1.115 20853
2 ...e_noMHC_cojo.txt 5826636 125935 125935 1.421 1.430 128573
3 ...y_noMHC_cojo.txt 5826636 405072 405072 1.688 1.695 409059
Estimated Omega: [[ 2.940e-05 7.324e-06 -2.653e-07] [ 7.324e-06 1.605e-05 -3.531e-07] [-2.653e-07 -3.531e-07 1.831e-06]] (Correlation): [[ 1. 0.337 -0.036] [ 0.337 1. -0.065] [-0.036 -0.065 1. ]] Estimated Sigma: [[1.014 0.068 0.016] [0.068 1.099 0.068] [0.016 0.068 0.918]] Best, Chen

[KKFFCC]

Hi,Patrick But this is not just a strand problem. After I used the "--incld_ambig_snps" parameter, the number of significant sites did not increase. Best, Chen

[paturley]

Hi Chen,

Many SNPs appear to be dropped during the MTAG QC process. If you just consider the set of overlapping SNPs between the GWAS and MTAG files, does it still look strange to you?

Patrick

On Thu, Dec 2, 2021 at 2:46 AM KKFFCC @.***> wrote:

Hi,Patrick But this is not just a strand problem. After I used the "--incld_ambig_snps" parameter, the number of significant sites did not increase. Best, Chen

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/JonJala/mtag/issues/148#issuecomment-984370665, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFBUB5J6VLOCZJ7G3ZZ4TS3UO4P65ANCNFSM5JDH5UNQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[KKFFCC]

Hi,Patrick Although there are more points that are filtered, the original significant points in the trait are still retained, but most of them become no longer significant after MTAG analysis.

Chen

[paturley]

Are you you using the Z-N option or the beta-SE option? We've noticed a few things that look like bugs in the beta-SE option recently, so it may be that the results using Z-N is more reliable.

On Thu, Dec 2, 2021 at 10:34 AM KKFFCC @.***> wrote:

Hi,Patrick Although there are more points that are filtered, the original significant points in the trait are still retained, but most of them become no longer significant after MTAG analysis.

Chen

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/JonJala/mtag/issues/148#issuecomment-984739374, or unsubscribe https://github.com/notifications/unsubscribe-auth/AFBUB5MD7QUYKXPGPVZ62ZLUO6G2HANCNFSM5JDH5UNQ . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

[KKFFCC]

I used the beta-se option. Thank you for your suggestion. I will experiment with the Z value to see if there is any improvement.