REF==ALT for variants with ALT as major

[cindywen96]

Hi! Thanks for building the tool.

I used crossmap to liftover VCF from hg38 to hg19. Some variants in the original VCF has REF as minor allele, ALT as major allele (the REF and ALT alleles agree with reference data like dbSNP. Just their relative frequencies are flipped in my data). I processed the data using plink1.9, which automatically put the major allele as A2 and when I converted it back to VCF, REF and ALT alleles got flipped. I next used crossmap to liftover the data. But these variants with flipped REF/ALT were unmapped with tag Fail(REF==ALT). I wonder why after liftover, REF is the same as ALT.

(I am aware of --no-comp-alleles flag, but I do want variants in the output data)

Thanks!

[liguowang]

"Fail(REF==ALT)" means at least one of the alleles in the "ALT" field is the same as the "REF" filed. If you use the " --no-comp-alleles" flag, such variants were still filtered out?

Liguo

On Fri, Apr 30, 2021 at 5:41 PM Cindy Wen @.***> wrote:

Hi! Thanks for building the tool.

I used crossmap to liftover VCF from hg38 to hg19. Some variants in the original VCF has REF as minor allele, ALT as major allele (the REF and ALT alleles agree with reference data like dbSNP. Just their relative frequencies are flipped in my data). I processed the data using plink1.9, which automatically put the major allele as A2 and when I converted it back to VCF, REF and ALT alleles got flipped. I next used crossmap to liftover the data. But these variants with flipped REF/ALT were unmapped with tag Fail(REF==ALT). I wonder why after liftover, REF is the same as ALT.

(I am aware of --no-comp-alleles flag, but I do want variants in the output data)

Thanks!

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/liguowang/CrossMap/issues/29, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACN443RWVZ6MCDG7UYPGMELTLMWZXANCNFSM435T4RHA .

[cindywen96]

For example, a SNP is G>A in reference, but A>G in my data. In my VCF, REF is A, ALT is G. Variants like these (flipped REF/ALT) were unmapped and the tag is Fail(REF==ALT). I wonder why this is so. I didn't use --no-comp-alleles. If I use it, the variant will be mapped? What will its REF and ALT be in the mapped file?