Fig. 1. Suitability of optimal target sets for drug repurposing.

A quantitative ranking of 30 model-predicted idealized interventions of various cardinality (green bars), length of response trajectory or efficiency (orange bars) and distance from optimal outcome (blue bars) as a function of their actionability (red line) or potential for translation into clinic using repurposing of currently available drugs. Only 16 of the initial 30 solution sets are significantly more accessible (empirical null distribution P < 0.05; A > A_{crit, P=0.05} ~7.25) to repurposing than a random MIS solution set (dashed blue line).