Fig. 1

Overlap graph of the top 25 enriched IPA-pathways. Each node represents one pathway significantly enriched with BMI-associated gene-specific transcripts. The node size is proportional to the number of BMI-associated transcripts within each respective pathway. Edge shade and width refer to the overlap between pathways in terms of shared transcripts and were calculated based on the Jaccard similarity coefficient, which is defined as the size of the intersection divided by the size of the union of two sample sets. Edges between nodes are shown if transcripts are shared between pathways and if the Jaccard coefficient ≥ 90th percentile. Nodes, i.e., pathways are colored according to the three defined gene expression signatures. The extensive connectivity, particularly for the attenuated insulin signaling, represents the strong overlap among the pathways. Pathways are numbered according to their enrichment p-values. 1: EIF2 Signaling, 2: Mitochondrial Dysfunction, 3: Regulation of eIF4 and p70S6K Signaling, 4: PI3K/AKT Signaling, 5: Insulin Receptor Signaling, 6: Hypoxia Signaling in the Cardiovascular System, 7: Chronic Myeloid Leukemia Signaling, 8: Production of Nitric Oxide and Reactive Oxygen Species in Macrophages, 9: Pancreatic Adenocarcinoma Signaling, 10: ILK Signaling, 11: AMPK Signaling, 12: JAK/Stat Signaling, 13: Renal Cell Carcinoma Signaling, 14: NRF2-mediated Oxidative Stress Response, 15: Ceramide Signaling, 16: mTOR Signaling, 17: CTLA4 Signaling in Cytotoxic T Lymphocytes, 18: B Cell Receptor Signaling, 19: Prostate Cancer Signaling, 20: Glioma Signaling, 21: Oxidative Phosphorylation, 22: Non-Small Cell Lung Cancer Signaling, 23: Heme Biosynthesis II, 24: Cyclins and Cell Cycle Regulation, 25: PDGF Signaling