Journal: Human Molecular Genetics

Article Title: Ancestry-specific high-risk gene variant profiling unmasks diabetes-associated genes

doi: 10.1093/hmg/ddac255

Figure Lengend Snippet: Network analysis reveals associations of candidate genes with UBC and with known T2DM-associated genes. (A) Unsupervised protein–protein interaction network analysis mapped 40 of the 57 candidate T2DM-associated genes identified in this study to a network centered on UBC. These 40 candidate genes (nodes highlighted in blue) include 11 previously associated with the T2DM phenotype (red font) and 29 genes with novel T2DM associations (gold font) (see also Table 2). The 31 genes within the dashed yellow circle have direct protein–protein interactions with UBC (central purple node). The nine genes outside of the dashed yellow circle indirectly interact with the UBC gene. The non-highlighted reddish-purple nodes represent other genes in the network. The protein–protein interaction network analysis utilized the OmicsNet (www.omicsnet.ca) bioinformatics platform with the InnateDB option. (B) Of the remaining 17 novel candidate genes (those unrelated to the UBC gene network), 16 genes (yellow font) could be mapped to one of two discrete networks using the IPA system (QIAGEN). Note that the PNLIPRP3 gene forms its own separate network (left side). Both networks could be additionally linked to a total of 111 known T2DM-associated genes (blue nodes, green font) using the ‘Grow’ function of IPA. The known T2DM-associated genes shown were from among a set of 4280 genes associated (P < 0.05) with the Type 2 Diabetes phenotype available from the T2DMKP. The grey nodes labeled with red font represent gene networks identified by IPA core analysis. The grey lines in both (A) and (B) represent interactions between different nodes. (See Supplementary Material, Table S8 for further details about the genes shown in the networks).

Article Snippet: The findings were further confirmed by clustered heat mapping of the hrVI across 20 428 genes for populations of different ancestry , using the commonly used bioinformatics software platform, Partek Genomic Suite.

Techniques: Protein-Protein interactions, Labeling