GWAS Meta analysis

Question

[alessandrotestori7]

Hi everybody. I am doing a sort of meta-analysis between two different GWAS datasets in order to see whether particular SNPs or sets of SNPs contribute to both pathological conditions. I am currently using METAL with the inverse variance-based method. However, when a SNP is very significantly associated to a disease and not to the other, the meta-analysis p-value is still very significant and this spoils the final result. What can I do to avoid such inconvenience? Please let me know. Thanks. Cheers!

[alessandrotestori7]

Any hint? Please let me know.