TyG Index Instrumental Variable SNP Dataset Based on UK Biobank

This dataset is based on publicly available genome-wide association studies (GWAS) data from the UK Biobank, with the core being single nucleotide polymorphism (SNP) instrumental variables associated with the triglyceride glucose (TyG) index. The original dataset was based on 192 TyG related SNPs screened from 273368 European populations (40-69 years old) (screening criteria: P<10 ⁻⁸, F>10 to avoid weak instrumental variable bias, r ²<0.01 and kb=10000 to remove linkage disequilibrium, and to exclude metabolic confounding sites such as BMI and blood pressure). After further quality control, Mendelian random multi-directional residual and outlier test (MR-PRESSO) was used to remove outlier SNPs (rs2081687, rs483082), and the allele information of the outcome variable (non-alcoholic fatty liver disease) GWAS data was matched. After removing palindromic structure SNPs, 139 valid SNPs were ultimately retained. The dataset provides detailed records of key genetic information such as effect allele (EA), other allele (OA), effect value (BETA), standard error (SE), and associated P-value for each SNP. It can be directly used for Mendelian randomization causal inference studies of TyG index and metabolic diseases (such as non-alcoholic fatty liver disease) in two samples, providing high-quality genetic tool variable support for etiological exploration of related diseases.