To assess SNPs in different genomic elements, we downloaded two SNP databases: (1) 12,891 SNPs from the National Human Genome Research Institute’s GWAS catalogue (https://www.genome.gov/26525384); (2) 14,416,369 common SNPs from dbSNP Build 142 common (downloaded from the UCSC Genome Browser) (treated as background variation). (PMCID: PMC5316832) I want to known where the 12,891 SNPs come from. My workflow is as follows: Step1: open the database: http://www.ebi.ac.uk/gwas/docs/file-downloads Step2: click on "All associations" and download. Step3: filter SNPs by p-value<1 × 10−5 (PMCID: PMC3965119) Step4: count the uniq SNPs However, the number of uniq SNPs is not 12,891. I want to known why and how? Thank you very much!