Research on LOH data of lung cancer using clustering algorithm

There exists close relationship between LOH phenomenon and malignant tumor. The approach to analyzing LOH data by clustering algorithm can find the SNPs loci related to cancer. Traditional clustering algorithms generally have some limitations, such as the sensitivity to initializing parameter, difficulty of finding out the optimized clustering results and the validity of clustering. In this paper, an efficient method for LOH analysis based on Artificial Fish Swarm Algorithm and k-means was proposed to improve the traditional algorithms. First, an Artificial Fish Swarm Algorithm was applied to the LOH data self-organized. As a result, an initial cluster center with the number of cluster of k-means was obtained; secondly, k-means was conducted to optimize the initial clustering result. The experimental results demonstrate the effectiveness and accuracy of our method in discovering chromosome segments related to suppressor genes of cancer.