Fig. 1

Overview of the data integration pipeline. We collected variants in the 1-Mb interval surrounding index variants at each of the 101 T2D GWAS loci along with relevant annotations for all protein coding genes in GENCODE including coding exon location, promoter location, distal regulatory elements correlated with gene activity from DNAseI hypersensitivity (DHS) data and summary statistic expression QTL (eQTL) data from T2D-relevant tissues. This, combined with information at the gene level from a semantic similarity metric, allowed us to define positional candidacy scores (PCS) for each gene in the GWAS intervals. PCS confers “weights” to the genes in the 1-Mb window based on biological candidacy, in contrast to “nearest” approaches, where the closest gene to the GWAS signal gets a weight of 1 and others a 0, or the “equal” approach where all genes in the window have the same weight. Genes with cumulative PCS > 0.7 were projected into the InWeb3 dataset using a Steiner tree algorithm to define a PPI network that maximises candidate gene connectivity. This network was further analysed to find processes, pathways and genes implicated in the T2D pathogenesis