Abstract
Prediction of drug-target interactions (DTI) plays a vital role in drug development in various areas, such as virtual screening, drug repurposing and identification of potential drug side effects. Despite extensive efforts have been invested in perfecting DTI prediction, existing methods still suffer from the high sparsity of DTI datasets and the cold start problem. Here, we develop KGE_NFM, a unified framework for DTI prediction by combining knowledge graph (KG) and recommendation system. This framework firstly learns a low-dimensional representation for various entities in the KG, and then integrates the multimodal information via neural factorization machine (NFM). KGE_NFM is evaluated under three realistic scenarios, and achieves accurate and robust predictions on four benchmark datasets, especially in the scenario of the cold start for proteins. Our results indicate that KGE_NFM provides valuable insight to integrate KG and recommendation system-based techniques into a unified framework for novel DTI discovery.
Highlights
Prediction of drug-target interactions (DTI) plays a vital role in drug development in various areas, such as virtual screening, drug repurposing and identification of potential drug side effects
We proposed a unified framework called KGE_NFM (Fig. 1) by incorporating knowledge graph embedding (KGE) and recommendation system techniques for DTI prediction that are applicable to various scenarios of drug discovery, especially when encountering new proteins
KGE_NFM, which could be viewed as a pre-trained model based on knowledge graph and is integrated with a recommendation system tailored for a specific downstream task, captures the latent information from heterogeneous networks using KGE without any similarity matrix and applies neural factorization machine (NFM) based on recommendation system to enforce the feature representation for a specific downstream task, which is the DTI prediction in this work
Summary
Prediction of drug-target interactions (DTI) plays a vital role in drug development in various areas, such as virtual screening, drug repurposing and identification of potential drug side effects. (2) Most recent methods are not evaluated in real-world scenarios in which one needs to make DTI prediction when new protein targets are identified for a complicated disease and elucidate molecular mechanisms of drugs with known therapeutic effects[39] This problem, similar to the cold start problem for recommendation systems, is a severe limiting factor for the practical application of DTI prediction methods. KGE_NFM, which could be viewed as a pre-trained model based on knowledge graph and is integrated with a recommendation system tailored for a specific downstream task, captures the latent information from heterogeneous networks using KGE without any similarity matrix and applies neural factorization machine (NFM) based on recommendation system to enforce the feature representation for a specific downstream task, which is the DTI prediction in this work. All of these results indicate that KGE_NFM is a powerful and robust framework with high extendibility for DTI prediction, which may provide new insights into the novel drug target discovery
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
