RESUMEN
Prediction of the interactions between small molecules and their targets play important roles in various applications of drug development, such as lead discovery, drug repurposing and elucidation of potential drug side effects. Therefore, a variety of machine learning-based models have been developed to predict these interactions. In this study, a model called auxiliary multi-task graph isomorphism network with uncertainty weighting (AMGU) was developed to predict the inhibitory activities of small molecules against 204 different kinases based on the multi-task Graph Isomorphism Network (MT-GIN) with the auxiliary learning and uncertainty weighting strategy. The calculation results illustrate that the AMGU model outperformed the descriptor-based models and state-of-the-art graph neural networks (GNN) models on the internal test set. Furthermore, it also exhibited much better performance on two external test sets, suggesting that the AMGU model has enhanced generalizability due to its great transfer learning capacity. Then, a naïve model-agnostic interpretable method for GNN called edges masking was devised to explain the underlying predictive mechanisms, and the consistency of the interpretability results for 5 typical epidermal growth factor receptor (EGFR) inhibitors with their structure‒activity relationships could be observed. Finally, a free online web server called KIP was developed to predict the kinome-wide polypharmacology effects of small molecules (http://cadd.zju.edu.cn/kip).
RESUMEN
Protein kinases are key components of cell signaling networks and thereby regulate fundamental biological processes such as cellular growth, proliferation, metabolism and survival. Kinome refers to all kinases in cells or tissue and "kinomics" is the global analysis of kinome with respect to abundance, activity, substrate specificity, phosphorylation pattern and mutational status. Human kinome currently contains 568 members, nearly half of which can be mapped to disease loci and deregulation of kinase activity by gene amplification or mutations has been implicated in diseases such as inflammation, diabetes and cancer. Therefore, human kinome is being recognized as a potentially rich source of drug targets. Kinase inhibitors have been successfully used to treat many kinds of advanced cancers. Chemical proteomics is emerging as a novel comprehensive kinome approach that combines an immobilized inhibitor affinity pull-down approach with mass spectrometry-based proteomics for kinase identification, quantification and phosphorylation analysis under physiological condition. Commonly, one or multiple broad-spectrum kinase inhibitors are covalently immobilized on a biocompatible matrix such as sepharose to enrich all kinases in cells or tissue and then the kinases are identified and quantified by mass spectrometry analysis. It can be used to study the specificity of kinase inhibitor drug, drug candidate or drug resistance mechanism, which can help to understand the mechanism and find combinational drug target. Large-scale unbiased kinome and cancer kinome study will facilitate new drug target discovery and correlate tumor tissue kinome profiles with response to therapy and therefore may be used for future therapy selection in personalized medicine. In this paper, the human kinome, kinase, kinase inhibitor and cancer, chemical proteomics based kinome study progress and its applications in drug discovery are reviewed.
RESUMEN
Protein kinases are key components of cell signaling networks and thereby regulate fundamental biological processes such as cellular growth, proliferation, metabolism and survival. Kinome refers to all kinases in cells or tissue and “kinomics”is the global analysis of kinome with respect to abundance, activity, substrate specificity, phosphorylation pattern and mutational status. Human kinome currently contains 568 members, nearly half of which can be mapped to disease loci and deregulation of kinase activity by gene amplifica-tion or mutations has been implicated in diseases such as inflammation, diabetes and cancer. Therefore, human kinome is being recognized as a potentially rich source of drug targets. Kinase inhibitors have been successfully used to treat many kinds of advanced cancers. Chemi-cal proteomics is emerging as a novel comprehensive kinome approach that combines an immobilized inhibitor affinity pull-down approach with mass spectrometry-based proteomics for kinase identification, quantification and phosphorylation analysis under physiological condi-tion. Commonly, one or multiple broad-spectrum kinase inhibitors are covalently immobilized on a biocompatible matrix such as sepharose to enrich all kinases in cells or tissue and then the kinases are identified and quantified by mass spectrometry analysis. It can be used to study the specificity of kinase inhibitor drug, drug candidate or drug resistance mechanism, which can help to understand the mechanism and find combinational drug target. Large-scale unbiased kinome and cancer kinome study will facilitate new drug target discovery and correlate tumor tissue kinome profiles with response to therapy and therefore may be used for future therapy selection in personalized medicine. In this paper, the human kinome, kinase, kinase inhibitor and cancer, chemical proteomics based kinome study progress and its applications in drug discovery are reviewed.
RESUMEN
Reversible protein phosphorylation by protein kinases and phosphatases is a common event in various cellular processes. The eukaryotic protein kinase superfamily, which is one of the largest superfamilies of eukaryotic proteins, plays several roles in cell signaling and diseases. We identified 482 eukaryotic protein kinases and 39 atypical protein kinases in the bovine genome, by searching publicly accessible genetic-sequence databases. Bovines have 512 putative protein kinases, each orthologous to a human kinase. Whereas orthologous kinase pairs are, on an average, 90.6 percent identical, orthologous kinase catalytic domain pairs are, on an average, 95.9 percent identical at the amino acid level. This bioinformatic study of bovine protein kinases provides a suitable framework for further characterization of their functional and structural properties.