南京理工大学学报（自然科学版）2019，Vol.43Issue(1)：1-12,12.DOI:10.14177/j.cnki.32-1397n.2019.43.01.001
蛋白质残基接触图预测
Protein residue-residue contact map prediction
摘要
Abstract
Proteins are large biomolecules,consisting of one or more amino acids residues. Proteins are the most import components in living cells and are involved in almost every living process. The functions of proteins are mostly determined by their structures. Accurately prediction of protein contact maps plays an important role in protein three-dimensional structure prediction. It has been one of the hottest topics in bioinformatics to predict contact map. The background knowledge and the great significance of protein contact map prediction are firstly introduced. After that,we summarize some representative methods for contact map prediction,including correlation-based methods,direct coupling analysis methods and their post-process strategies. Supervised machine learning-based methods are also introduced in this section. Analysis and comparisons are made based on the performances of the most advanced methods in critical assessment of protein structure prediction ( CASP) competition. Applications in protein 3D modeling based on predicted contact map are also introduced. Finally,promising directions are provided to the key issues in contact map prediction.关键词
蛋白质残基接触图/蛋白质结构预测/协同进化/机器学习/国际蛋白质结构预测竞赛Key words
protein contact map/protein structure prediction/co-evolution/machine learning/critical assessment of protein structure prediction分类
信息技术与安全科学引用本文复制引用
於东军,李阳..蛋白质残基接触图预测[J].南京理工大学学报（自然科学版）,2019,43(1):1-12,12.基金项目
国家自然科学基金(61772273,61373062,61876072) （61772273,61373062,61876072）
