Protein-protein interaction sites prediction by ensemble random forests with synthetic minority oversampling technique
Protein-protein interaction sites prediction by ensemble random forests with synthetic minority oversampling technique
复制标题
使用合成少数过采样技术的集合随机森林预测蛋白质-蛋白质相互作用位点
DOI：
10.1093/bioinformatics/bty995
复制
发表时间：
2019-07-15
期刊：
影响因子：
5.8
通讯作者：
Ma, Qin
中科院分区：
文献类型：
--
作者：
Wang, Xiaoying;Yu, Bin;Ma, Qin
激励蛋白质 - 蛋白质相互作用（PPI）位点的预测是突变设计，催化反应和PPI网络重建的关键。考虑到样本中的重要序列和不平衡问题，这是一项具有挑战性的任务。提出了一种新的基于整体学习的方法，一种用于不平衡样本的合成少数群体过采样技术（SMOTE）的合奏学习，并提出了RF算法（EL-SMURF）（EL-SMURF）对于PPI站点的预测。使用滑动窗口将序列曲线特征和残基演化速率合并，以提取相邻残基的特征，并将SMOTE应用于特征空间中的超级界面残基，以解决不平衡问题。实现了多维缩放特征选择方法，以减少特征冗余和子集选择。最后，将随机的森林分类器应用于建立集成学习模型，并将最佳特征向量插入El-Smurf以预测PPI位点。 El-Smurf在两个独立验证数据集上的性能验证显示77.1％和77.7％的准确性，比其他现有工具分别高6.2-15.7％和6.1-18.9％。可利用和实施中使用的源代码和数据这项研究可在http://github.com/qust-aibbdrc/el-smurf/.supplextary上公开获得。在线生物信息学可以获得信息支持数据。
Motivation
The prediction of protein-protein interaction (PPI) sites is a key to mutation design, catalytic reaction, and the reconstruction of PPI networks. It is a challenging task considering the significant abundant sequences and the imbalance issue in samples.
Results
A new ensemble learning based method, EL-SMURF, was proposed for PPI sites prediction in this study. The sequence profile feature and the residue evolution rates were combined for feature extraction of neighboring residues using a sliding window, and the synthetic minority oversampling technique was applied to oversample interface residues in the feature space for the imbalance problem. The multidimensional scaling feature selection method was implemented to reduce feature redundancy and subset selection. Finally, the Random Forest classifiers were applied to build the ensemble learning model, and the best feature vectors were inserted into EL-SMURF to predict PPI sites. The performance validation of EL-SMURF on two PPI datasets showed 77.1% and 77.7% accuracy, which were 6.2%-15.7% and 6.1%-18.9% higher than the other existing tools, respectively.
Availability
The source codes and data used in this study are publicly available at http://github.com/QUST-AIBBDRC/EL-SMURF/.
Supplementary information
Supplementary data are available at Bioinformatics online.