Additional supplementary text and explanation of methodology, together with supplementary data tables and figures referred to in the main document.
The full annotation and network-based druggability predictions of the 13,345 proteins in this analysis (using the largest interactome). The full prediction results for 10,998 proteins using the largest Y2H-based models; model quality and AUCs for the Y2H models.
Details of the top most druggable proteins identified using a network-based druggability analysis that are not themselves targets of FDA-approved drugs
Individual predictive results and relative information content of each of the topological, community and graphical features used to train the models.
File containing the raw data used to generate the correlation plot in Figure F in S1 Text showing the limited correlation observed between the network topological, graphical and community-based features used in our analysis.
P-values of association between individual features of the drug target classes, namely all drug targets, targets of cancer drugs, or targets of drugs used in other therapeutic areas.
canSAR: updated cancer research and drug discovery knowledgebase
Krishna C. Bulusu, Joseph E. Tym, Elizabeth A. Coker, Amanda C. Schierz, and Bissan Al-Lazikani.
Nucleic Acids Research 2014; doi: 10.1093/nar/gkt1182