I am doing GWAS meta-analysis, and met a new issue that I never had before. I have 2 datasets that had different genotyping array. I used 1000 G phase 3 reference panel to impute both datasets. When meta-analyze in METAL, some of the SNPs that have different rsid (for example, rs1 and rs1:123:G:A) in two datasets are only shown in one study and missing in another. How can I make all same SNPs have same rsid expression in the datasets with millions of SNPs?