Auto-layout of ‘Caffeine (C) neighborhood’ based on Ingenuity curated finding (IPA, Ingenuity® Systems)
Auto-layout of ‘Caffeine (C) neighborhood’ based on Ingenuity curated finding (IPA, Ingenuity® Systems). The nature of interaction although available in the database or incorporated in labels of edges/arcs, was stripped from graphs to reduce the clutter, save for an example of maximal motif NF-κB-TAC1-TNF isolated on the right (see text for explanations). In the latter case, and in the rest of the nodes, self-connections were removed. The type of interaction is only identified as either ‘direct’ or ‘indirect’ (dashed lines stand for indirect interactions; solid lines indicate direct interactions). The various shapes of nodes denote the functional class of the molecules that are easily understood from their abbreviations. In order to reduce the clutter, molecular interactions that appear unrelated to the effect were trimmed. Additional molecules of interest were simply inserted by entering their terms into IPA search engine one by one or using multiple terms searches and endorsing all desirable information on connections in the ‘New pathway’ (or any available stored pathway) window via commands, such as ‘connect,’ or ‘path analysis’ based on PubMed data. Path analysis is based on network property known as node ‘betweenness’; the number of shortest paths between any chosen nodes. The latter is the smallest number of steps between any arbitrarily chosen nodes in the network selected for analysis. When there were several shortest paths between the nodes, the first one was selected as better substantiated by literature. These commands permitted to manually edit networks and in this way to create custom molecular pathways supported by the current experimental literature evidence. Color-coding depicts multiple roles that molecules may play in several signaling pathways. Arbitrarily selected are: Apoptosis of endothelial cells (p=1.41E-12); ROS production (p=5.11E-11); Energy expenditure (p=4.16E-14); and Thermoregulation (p=2.34E-15. More than four functions would make the graph more confusing in the chosen format. The probability that each biological function assigned to the set of molecules is due to chance alone was determined using Fischer’s exact test.