search

dajiangliu / rareGWAMA

5 stars 6 forks source link

Flipped REF/ALT alleles #1

Closed sarahgra closed 4 years ago

sarahgra commented 4 years ago

I'm trying to do a conditional analysis using rvtests generated input files. In some studies for a given chromosome and position, the REF/ALT might be C/T while others may have it coded oppositely as REF/ALT T/C. It seems that rareGWAMA treats these as two different variants rather than being equivalent (whereas RAREMETAL and METAL handle the flipping internally). Is there any way of having rareGWAMA handle this internally rather than having to reformat the summary statistics files?

dajiangliu commented 4 years ago

Yes, for rareGWAMA, it mostly deals with imputation based GWAS, so I assumed that the strand and ref/alt is correct. For rareMETALS, it was developed for exome chip studies, where the allele labels are incorrect all the time, so we have careful checks there (but it is slow) If the flip in your study is not too many, would you please reformatted manually? I can try to implement it as well to give user an option to check flips.

All the best, Dajiang

Associate Professor Dept. Public Health Sciences Institute of Personalized Medicine Penn State College of Medicine, HCAR 2020 Email: dajiang.liu@psu.edu URL: https://dajiangliu.blog https://dajiangliu.blog Tel: +1-717-531-4178

On Wed, Apr 29, 2020 at 10:59 AM sarahgra notifications@github.com wrote:

I'm trying to do a conditional analysis using rvtests generated input files. In some studies for a given chromosome and position, the REF/ALT might be C/T while others may have it coded oppositely as REF/ALT T/C. It seems that rareGWAMA treats these as two different variants rather than being equivalent (whereas RAREMETAL and METAL handle the flipping internally). Is there any way of having rareGWAMA handle this internally rather than having to reformat the summary statistics files?

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/dajiangliu/rareGWAMA/issues/1, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACNCDJML32CW2RWIDJKTFRDRPA6EBANCNFSM4MTZOI7Q .

sarahgra commented 4 years ago

Ok, I can probably reformat the study files then. These are imputation-based studies but are a mix of 1000 genomes, HRC, and custom imputation panels so some studies have the effect alleles coded differently.

dajiangliu commented 4 years ago

Thank you! BTW, we always report alternative allele effects as default.

All the best, Dajiang

Associate Professor Dept. Public Health Sciences Institute of Personalized Medicine Penn State College of Medicine, HCAR 2020 Email: dajiang.liu@psu.edu URL: https://dajiangliu.blog https://dajiangliu.blog Tel: +1-717-531-4178

On Wed, Apr 29, 2020 at 11:16 AM sarahgra notifications@github.com wrote:

Ok, I can probably reformat the study files then. These are imputation-based studies but are a mix of 1000 genomes, HRC, and custom imputation panels so some studies have the effect alleles coded differently.

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/dajiangliu/rareGWAMA/issues/1#issuecomment-621278360, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACNCDJKZRDGWLKPXCNTJHALRPBADLANCNFSM4MTZOI7Q .