ABSTRACT
This chapter introduces a computational approach, namely, neighborhood regularized logistic matrix factorization (NRLMF), to predicting potential interactions between drugs and targets. The novelty of NRLMF comes from integrating logistic matrix factorization with neighborhood regularization to predict the interaction probability of a given drug-target pair. Specifically, both drugs and targets are mapped into a shared latent space, and the drug–target interactions (DTIs) are modeled using the linear combinations of the drug-specific and target-specific latent vectors. The drug similarities are calculated based on the chemical structures of compounds, and the target similarities are computed based on the amino acid sequences of the target proteins. However, there are various types of side information associated with drugs and targets. The chapter discusses how the neighborhood information benefits DTI prediction under cross-validation setting (CVS). It reviews existing DTI prediction methods from two aspects: traditional classification-based methods and matrix factorization-based methods.
