Distributed ICSA Clustering Approach for Large Scale Protein Sequences and Cancer Diagnosis

OBJECTIVE: With the over saturating growth of biological sequence databases, handling of these amounts of data has increasingly become a problem. Clustering has become one of the principal research objectives in structural and functional genomics. However, exact clustering algorithms, such as partitioned and hierarchical clustering, scale relatively poorly in terms of run time and memory usage with large sets of sequences. METHODS: From these performance limits, heuristic optimizations such as Cuckoo Search Algorithm with genetic operators (ICSA) algorithm have been implemented in distributed computing environment. The proposed ICSA, a global optimized algorithm that can cluster large numbers of protein sequences by running on distributed computing hardware. RESULTS: It allocates both memory and computing resources efficiently. Compare with the latest research results, our method requires only 15% of the execution time and obtains even higher quality information of protein sequence. CONCLUSION: From the experimental analysis, We noticed that the cluster of large protein sequence data sets using ICSA technique instead of only alignment methods reduce extremely the execution time and improve the efficiency of this important task in molecular biology. Moreover, the new era of proteomics is providing us with extensive knowledge of mutations and other alterations in cancer study.