Mechanism Analysis of Antitumor Drugs based on MTDNN

Protein kinases are the main dependent targets of oncogenes, and antitumor drugs mainly target kinases. In order to analyze the mechanism of antitumor drugs, a small molecule kinase prediction model was constructed based on multi task deep neural network to predict the interaction between small molecule kinase inhibitors and kinases. The experimental results show that the small molecule kinase prediction model based on MTDNN has good prediction ability. The average auroc on different test sets is 0.7425. It can effectively predict and analyze the therapeutic targeting activities of various kinases and deeply analyze the action mechanism of antitumor drugs. The introduction of artificial intelligence technology into the mechanism research of antitumor drugs provides a reference for the research and development of accurate and targeted new antitumor drugs, and has important practical value for improving the efficiency and quality in the field of drug research and development.