Constructing a compound mode-of-action network for personalized drug effectiveness prediction

The problem of large scale computational drug screening is considered in this thesis. A new concept of Mode-of-Action (MoA) network (MoNet), or MoNet, is introduced to model the relationship of therapeutic effectiveness between different drugs. A new algorithm for constructing Mode-of-Action groups and subsequently MoNet based on microarray profile of drug treatment is proposed. The algorithm has been applied to the data from the Connectivity Map (cMap) project and a cMap MoNet was obtained. A new drug effectiveness prediction algorithm based on MoNet is also developed and applied to an independent breast cancer data set. The prediction result shows significant improvement in precision over cMap.