快速的基于蚁群聚类的PPI网络功能模块检测方法

针对蚁群聚类在蛋白质相互作用( protein-protein interaction,PPI)网络中进行功能模块检测问题上时间性能的不足,提出一种快速的基于蚁群聚类的PPI网络功能模块检测( fast ant colony clustering for functional module detection, FACC-FMD)方法。该算法计算每个蛋白质与核心组蛋白质的相似度,根据拾起放下模型进行聚类,得到的初始聚类结果中功能模块之间相似度很小,省去了原始蚁群聚类算法中的合并和过滤操作,缩短了求解时间。同时该算法根据蛋白质的关键性对蚁群聚类中的拾起放下操作做了更严格的约束,以减少拾起放下的次数,加速了聚类的过程。在多个PPI网络上的实验表明：与原始蚁群聚类方法相比,FACC-FMD大幅度提高了时间性能,同时取得了良好的检测质量,而且与近年来的一些经典算法相比在多项性能指标上也具有一定的优势。

Fast Ant Colony Clustering for Functional Module Detection Algorithm in PPI Networks

The time performance of ant colony clustering seriously restricts its application for functional module. A fast ant colony clustering for functional module detection ( FACC-FMD ) algorithm, which considerably speeded up the original ACC-FMD algorithm was developed. The similarity between each protein and core protein group was computed by the FACC-FMD, then clustered by the pick-up and drop-down model. The similarity between the functional modules by clustering was small. Thus FACC-FMD eliminated the need for the merge operation and filter operation in ant colony cluster, and shorten the running time. At the same time, the essential of protein was computed and was used to constraint the times of pick-up and drop-down. Experiments on multiple PPI networks show that the FACC-FMD algorithm can greatly improve the time performance of ant colony clustering for functional module detection with satisfactory quality. Moreover, compared with classical algorithms in recent years, the FACC-FMD also has advantages in performance indicators.