Prediction of drug-related phenotypes based on the constructed phenotype-gene-drug heterogeneous network

Understanding the relationship between disease phenotypes and drugs is an important issue in network medicine. With the continuous enrichment of phenotypic and drug data, scientists have proposed various approaches to explore biological information from these data. Among the existing methods, network-based ones show their advantages. However, most of the considered networks were centered on disease networks, and rarely considered networks that were composed of three or more biological entities. By integrating disease phenotypes, genes and drugs data that were extracted from three databases, a heterogeneous network that containing relationships among the three biological entities was constructed. Based on the constructed networks, we develop a new method called Relationship based on the Shortest Path (RSP) to identify disease phenotypes and drug relationships. In the RSP, we first calculated the global correlation between the 22 phenotype classes and the 14 drug classes, then the shortest path length between each disease phenotype and each drug in the heterogeneous network was used to improve such global correlation, and finally, we obtained the local correlation coefficients between drugs and phenotypes. Different indicators were used to evaluate the method, the RSP method outperforms the ProphNet and the Shortest Path (SP) methods in inferring potential phenotypic targets for drugs. Our investigations have potential implications in network medicine.