A chronopharmacology-friendly multi-target therapeutics based on AI: the example of therapeutic hypothermia

Abstract

Background Computer-aided drug discovery (CADD) is a widely used method for drug discovery with many successes. Meanwhile, CADD has the limitation of analyzing multi-level scores such as docking results of multiple proteins with multiple drugs. Results We propose a method of PageRank to solve the problem. This method can make a comprehensive ranking based on multi-level scores. Then we take an example of therapeutic hypothermia (TH). Three levels of TH data were used in the article: the log2 foldchange (logFC) of proteins, the relative expression values of mRNA, and the docking scores of proteins and molecules. After calculation, we get the comprehensive drug rank and drug combination rank of each group of TH, which means we can generate the rank of drug directly from bioinformatics. Based on this method, we raised the concept of bioinfo-pharmacology. Conclusions Given the high rationality and compatibility of bioinfo-pharmacology, it can effectively enhance popular drug discovery techniques such as the docking or pharmacophore model. Besides, it could advance the application of precision medicine.