Identifying Lysine Succinylation Sites in Proteins by Broad Learning System and Optimizing Imbalanced Training Dataset via Randomly Labeling Samples

• 作者: JIA Jianhua SHEN Yanxia QIU Wangren
• 来源期刊: 武汉大学自然科学学报（英文版）
• 年，卷（期）: 2021,（1）
• 所属期刊栏目: Computer Science
• 页码范围: 81-88
• 页数: 8页
• 分类号: TP305
• 语种: 英文
• DOI: 10.19823/j.cnki.1007-1202.2021.0005

摘要

As one important type of post-translational modifica-tions(PTMs),protein lysine succinylation regulates many impor-tant biological processes.It is also closely involved with some major diseases in the aspects of Cardiometabolic,liver metabolic,nervous system and so on.Therefore,it is imperative to predict the succinylation sites in proteins for both basic research and drug development.In this paper,a novel predictor called iSuccLys-BLS was proposed by not only introducing a new machine learning algorithm—Broad Learning System,but also optimizing the im-balanced data by randomly labeling samples.Rigorous cross-validation and independent test indicate that the success rate of iSuccLys-BLS for positive samples is overwhelmingly higher than its counterparts.