论文部分内容阅读
由于蛋白质的相互作用是动态变化的,因此使用常规检测方法从静态PPI网络数据中识别蛋白质复合物具有一定的局限性.本文结合时序基因表达数据,提出了一个基于时序功能保持特征和蚁群聚类的复合物检测算法.算法首先根据相邻时刻的子网结构,选出在相邻时刻都具有表达活性的种子节点集合.然后结合复合物的保持特征,构建一组与前一时刻复合物集合具有功能相似性的初始蛋白质簇集合,并利用蚁群聚类的拾起、放下规则,完成对其他蛋白质的聚类,从而形成最终的复合物.实验结果表明使用时序功能保持特征可以提高复合物预测的准确性,与其他方法相比,新算法在精度方面也具有较好的性能.
Due to the dynamic changes of protein interactions, the use of routine detection methods to identify protein complexes from static PPI network data has some limitations.In this paper, we propose a time-based functional retention and ant colony clustering Class complex detection algorithm.At the beginning of the algorithm, we select the set of seed nodes that have the activity of expression at the adjacent moment according to the sub-network structure of the adjacent moment.And then combine with the keeping characteristics of the complex, A set of initial protein clusters with functional similarities is assembled and the ant colony clustering is used to pick up and put down the rules to complete the clustering of other proteins to form the final complex.The experimental results show that the use of timing functions to maintain the characteristics can improve the composite Compared with other methods, the new algorithm also has better performance in accuracy.