The work of Yanan Tian, a student in the Joint Doctoral Programme between the University of Coimbra Faculty of Science and Technology and Macao Polytechnic University, promises a new approach to the design of drugs with greater selectivity and clinical efficacy.
“This work proposes the MMCLKin model, a structure based on advanced AI methods, designed to predict with high accuracy and interpretability the activity and selectivity of kinase inhibitors, significantly accelerating the process of discovering and optimising new targeted drugs,” explained Yanan Tian, in a statement from the Faculty of Science and Technology sent to the Lusa agency.
According to the note, protein kinases are one of the most relevant classes of therapeutic targets in biomedical research, whose potential results from the central role they play in regulating multiple cellular processes, including proliferation, differentiation, and cell death.
“However, the development of highly selective inhibitors remains a challenge due to the strong structural conservation between kinases and the high cost of experimental trials,” it said.
The results of the research carried out under the guidance of Professors Joel P. Arrais, from the Department of Computer Engineering at the University of Coimbra, and Huanxiang Liu, from Macao Polytechnic University, “demonstrate that the model outperforms existing methods in predicting the affinity and selectivity of inhibitors, even in cases involving unknown structures or mutated kinases.”
According to the study’s authors, the tests validated the model’s predictive power, demonstrating that five compounds suggested by MMCLKin effectively inhibit a mutation associated with neurodegenerative diseases, four of which are active in nanomolar concentrations.
“These results reinforce the potential of MMCLKin as a tool to accelerate the development of targeted therapies, opening up new perspectives for the rational design of drugs with greater selectivity and clinical efficacy,” the statement reads.
According to the researchers, the proposed approach represents a breakthrough in the application of AI to drug discovery, demonstrating how next-generation computational models can reproduce complex biological processes “in silico” that, experimentally, can take years or even decades.
This capability enables the rapid identification of promising therapeutic candidates, significantly reducing the time and cost of pharmaceutical research.
Beyond its immediate impact, it opens up new directions for research in the field of kinase modelling by providing a unified framework for analysing structural, mutational and functional patterns across the entire family of human kinases.
“This type of approach could evolve into generalist models capable of anticipating the behaviour of new kinases — including those without an identified structure — and support the rational design of selective and personalised therapies in various areas, from cancer to neurodegenerative diseases.”
Platform with Lusa
