蛋白质-配体绑定区域预测的特征抽取及学习算法研究

Protein-ligand interactions are ubiquitous and play important roles in a wide variety of biological processes. Accurately identifying the protein-ligand binding sites or pockets is of significant importance for understanding the protein function, analyzing the relationship between biological molecules, and instructing drug design etc. With the rapid development of protein sequencing technology, huge volumes of protein complexes have been generated in the post-genomic era and will continue to grow exponentially. It is highly desired to develop intelligent automatic methods for protein-ligand binding site prediction. In this study, we will systematically carry out research on feature extraction level and decision model level, respectively, for protein-ligand prediction based on the thorough evaluation on the shortcomings and bottlenecks of the existing prediction models. Contents of the research consist of protein evolution image sparse representation, protein multi-view feature dissimilarity analysis and extraction, ligand-specific prediction model, incremental prediction model, and interpretable prediction model. We expect to research and develop effective machine learning theory and algorithms for meeting the urgent demands of protein data analysis and processing from proteomics. Also, we will provide theoretically-sound, practically-good and highly interpretable protein computational models for biologists and general users. Our research will enrich the contents of protein computation, speed up the process of unveiling the relationship between protein sequence, structure and function, and further accelerate the development of bioinformatics.

蛋白质与其他生物分子(配体)之间的互相作用在生命过程中普遍存在并且不可或缺。准确识别出蛋白质与配体之间的相互作用区域，对于理解蛋白质的功能、分析生物分子之间的相互关系、指导药物设计等具有重要的指导意义。随着蛋白质数据的海量呈现，研发高效可靠的智能方法来进行蛋白质-配体绑定区域预测成为迫切需要。本项目在分析现有预测模型存在不足和瓶颈的基础上，从特征抽取和模型学习算法两个层面展开系统研究，以进一步提高蛋白质-配体绑定预测方法的可靠性和可用性。具体研究内容包括：蛋白质进化图象的稀疏特征抽取、蛋白质多视角特征差异性分析及抽取、配体特异性预测模型、增量式预测模型以及可解释性预测模型。本研究旨在为生物学及相关研究人员提供理论依据充分、实用性强、具有良好可解释性的应用系统，亦是对蛋白质生物计算领域的丰富与发展，有助于进一步推动生物信息学的发展。

蛋白质通过绑定位点、区域等形式与配体之间产生相互作用，这些相互作用在生命活动中是广泛存在且不可或缺的。精确地识别蛋白质与配体之间的绑定位点及区域，对于理解蛋白质的功能、分析生物分子之间的相互关系、药物设计等具有重要的指导意义。随着蛋白质数据的海量呈现及持续增长，蛋白质大数据时代已然到来。在大数据时代下，研发智能高效的蛋白质-配体绑定区域预测方法是十分有必要的。.本项目完成了基于多视角特征学习的配体绑定位点模型、基于类不平衡学习的配体绑定位点模型、基于海量蛋白质数据分析的配体绑定位点预测模型以及基于多标签学习算法的配体绑定位点预测模型的学习与构建；发表相关SCI/SCIE期刊论文20篇、国际会议论文2篇；授权国家发明专利4项及软件著作权4项。上述研究成果进一步地提高蛋白质-配体绑定预测方法的可靠性和实用性。为了便于实际应用，本项目为生物学及相关研究人员提供了理论依据充分、实用性强、交互性良好的应用服务系统。本项目是对蛋白质生物计算领域的丰富，有助于进一步推动生物信息学的发展。