Experimental determination of Drug-Target Interactions (DTIs) is expensive and time-consuming. Hence, there is a continuous demand for more accurate DTI predictions using computational techniques. In this work, we propose a Dual-Network Integrated Logistic Matrix Factorization (DNILMF) algorithm to infer potential DTIs. We compare DNILMF to the prior art based on the benchmark dataset. Results indicate DNILMF outperforms its counterpart in terms of the recall based evaluation metrics, such as Mean Percentile Ranking (MPR), in the case of 5 trials of 10-fold cross-validation. We hope the current work can increase research productivity toward drug repositioning and polypharmacology.<br/>

Predicting drug-target interactions by dual-network integrated logistic matrix factorization

254th National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.