Many mismatch allele for meta analysis by plink

Question

[khn]

Hello,

I am doing meta analysis by plink. Given many allele mismatch, no SNPs in ChrX remained after meta analysis, though it went well for Chr1-22. I figured out that the issues are flip.

I do not think it is correct just to flip A1 and A2 for SNPs in ChrX in one summary statistics, and do meta analysis again for the two summary statistics.

Thus, does anybody know how to treat this?

Thank you in advance!