Method and system for predicting drug-target protein interaction relationship based on decision template

Abstract

The invention discloses a method and a system for predicting a drug-target protein interaction relationship based on a decision template. Multiple similarity characteristic combinations are formed in combination with existing drug compound molecular structure similarity, drug ATC annotation similarity, target protein sequence similarity and function similarity through proposing two new target protein similarity measurement policies, namely, GO ontology annotation-based and pathway function mapping-based similarity measurement; and the drug-target protein interaction relationship is predicted by adopting a KNN classification algorithm based on a hypothesis that similar drugs easily interact with similar target proteins. According to the method and the system, a decision template fusion-based policy is proposed; multiple similarity measurement-based classifier prediction results are subjected to decision level fusion; the problem that a known drug-target protein interaction relationship is relatively sparse is effectively solved in combination with concepts of super target proteomes and super drug groups; the prediction precision is improved; and the method and the system can be used for realizing target protein prediction of new drugs or drug prediction of new target proteins.

Description

technical field [0001] The invention relates to the fields of biology and information technology, in particular to a method and system for predicting drug-target protein interaction relationship based on a decision template. Background technique [0002] Developing a new drug generally costs billions of dollars, takes 9-12 years, and there is a risk of high failure rate and high recall rate. According to statistics, since 1950, the number of newly authorized drugs is almost zero (reference Literature: Scannell JW, Blancley A, Boldon H, et al. Diagnosing the decline in pharmaceutical R&Defficiency [J]. Nature reviews Drug discovery, 2012, 11(3): 191-200.). However, with the development of sequencing technology and biotechnology, a large amount of drug and biological omics data have been generated. Studying new target proteins of existing drugs can be used to discover new uses of existing drugs, and at the same time, it can also discover the off-target of the drug. The forme...

AI Technical Summary

Problems solved by technology

[0004] Traditional ligand-based methods and target protein-based methods require the three-dimensional structure of the target protein to be known, and can only predict target proteins with known drugs

In addition, most existing methods only make predictions based on the structural similarity of drugs and the sequence similarity of target proteins, which cannot reflect the fact that drugs with different structures may interact with the same target protein and targets with different sequence similarities. The protein may interact with the same drug, and for the proposed multiple similarity measures, most of the combinations are averaged and maximized, which is too simple to make full use of the proposed multiple similarities. Sexual measurement (references: Y.Yamanishi, M.Kotera, Y.Moriya, R.Sawada, M.Kanehisa, S.Goto, (2014) DINIES:drug–target interaction networkinference engine based on supervised analysis, Nucleic acids research, 42W39-W45.J.-Y.Shi, S.-M.Yiu, Y.Li, H.C.Leung, F.Y.Chin, (2015) Predicting drug–target interaction for new drugs using enhanced similarity measures and super-target clustering, Methods, 83 98-104.)

Images