AUTHOR=Lee Yeeun , Eun Jisu , Lee Jinhyuk , Nam Seungyoon 

TITLE=DeepKinome: quantitative prediction of kinase binding affinity by a compound using deep learning based regression model

JOURNAL=Frontiers in Molecular Biosciences

VOLUME=Volume 12 - 2025

YEAR=2025

URL=https://www.frontiersin.org/journals/molecular-biosciences/articles/10.3389/fmolb.2025.1698891

DOI=10.3389/fmolb.2025.1698891

ISSN=2296-889X

ABSTRACT=IntroductionKinases are essential for cellular regulation and drug development. Predicting the quantitative binding affinity between small-molecule compounds and kinases remains a challenge because of data complexity.MethodWe developed DeepKinome, a 20-layer convolutional neural network-based deep learning (DL) regression model, to predict quantitative binding affinity. Given the continuous nature of binding affinity, the root mean square error (RMSE), the coefficient of determination (R2), the Pearson’s correlation coefficient (PCC) between actual and predicted values, and the acceptance interval ratio (AIR) were evaluated. Trained using data from 234 kinases and 163 compounds from the L1000 database.ResultsDeepKinome outperformed five DL and four machine learning models, achieving an RMSE of 1.157, an R2 of 0.535, a PCC of 0.743, and an AIR of 0.570. Explainable artificial intelligence analysis revealed key amino acid sequences that influenced the predictions aligned with known kinase phosphorylation sites.ConclusionDeepKinome offers a promising approach for understanding kinase inhibition and compound binding.