Chunk #42 — 6. The wider picture of replication efforts: consortia, data availability, and field synopses — 6.i. Example from the field of type 2 diabetes
Researchers in the field of type 2 diabetes (T2D) genetics were among the first to lead the way in distributed collaborative networks, exemplified by early efforts such as the International Type 2 Diabetes Linkage Analysis Consortium and the International Type 2 Diabetes 1q Consortium. [66–68] The advent of GWA scans was met by pre-publication data sharing between three large-scale studies, the WTCCC, DGI and FUSION scans, [9, 69–71] leading to the formation of the DIAGRAM Consortium (Diabetes Genetics Replication and Meta-analysis). By exchanging information on top signals, the three studies obtained in silico replication of individual scan findings and then further pursued de novo replication in additional sets of independent samples. This endeavour additionally highlighted examples of statistical heterogeneity across the studies, notably with respect to one of the WTCCC study’s strongest signals, residing within the FTO gene. [72] This inconsistency in observed associations could be ascribed to study design and, specifically, to matching cases and controls for BMI (DGI study). The FTO signal was quickly identified as the first robustly replicating association with obesity, mediating its effect on T2D
