Quantifying Cancer Drug Targetability across Biochemical Pathways

The evaluation of cancer drug treatments within the complex context of biological networks and pathways is challenging. Drugs can be "promiscuous," targeting multiple genes or pathways, not just the ones of interest. This complexity makes it difficult to assess the full impact of a drug solely based on individual knowledgebases. We make the impact more comprehensible through the characterization of drug targetability in terms of a combination of biochemical pathway data from the Reactome Knowledgebase and drug-target interaction data from the Cancer Targetome. This leads to a graphical representation of drug targetability. We further improve comprehensibility by developing metrics to quantify the graph structures in order to evaluate drug target specificity and navigate the hierarchical nature of biological networks and identify candidate treatment options to increase specificity and minimize off-target effects. We have developed software that calculates these metrics for drugs in the Cancer Targetome. The software is available for research purposes only at https://github.com/sehagler/drug_targetability.