Disentangling the consequences of type 2 diabetes on targeted metabolite profiles using causal inference and interaction QTL analyses

• DOI: 10.1371/journal.pgen.1011346
• Publication Date: 2024-12-03T18:38:04Z
• Authors: Ozvan Bocher; Archit Singh; Yue Huang; Urmo Võsa; Ene Reimann; Ana Arruda; Andrei Barysenska; Anastassia Kolde; Nigel W. Rayner; ; Tõnu Esko; Reedik Mägi; Eleftheria Zeggini
• Keywords: Mendelian Randomization; Genome-wide Association Study; Genetic Association

Circulating metabolite levels have been associated with type 2 diabetes (T2D), but the extent to which T2D affects and their genetic regulation remains be elucidated. In this study, we investigate interplay between genetics, metabolomics, risk in UK Biobank dataset using Nightingale panel composed of 249 metabolites, 92% correspond lipids (HDL, IDL, LDL, VLDL) lipoproteins. By integrating these data large-scale GWAS from DIAMANTE meta-analysis through Mendelian randomization analyses, find 79 metabolites a causal association T2D, all spanning lipid-related classes except for Glucose Tyrosine. Twice as many are causally affected by liability, almost tested classes, including branched-chain amino acids. Secondly, an interaction quantitative trait locus (QTL) analysis, describe four consistently replicated independent Estonian Biobank, loci two different genomic regions show attenuated cases compared controls. The significant variants QTL analysis QTLs corresponding general population not risk, pointing towards consequences on levels. Finally, differential level 165 microvascular, macrovascular, or both types complications, only few discriminating complication classes. Of 40 linked either direction, suggesting biological mechanisms specific occurrence complications. Overall, work provides map targeted regulation, enabling better understanding trajectory leading