AN EFFECTIVE REVERSE ENGINEERING METHODOLOGY FOR THE RECONSTRUCTION OF GENE REGULATORY NETWORK USING GENE EXPRESSION PROFILE

Gene regulatory network reconstruction using microarray data is a challenging problem in bioinformatics. Recognition of relationships between various genes in a biological process has great impact on the various application areas such as drug design and treatment of genetic disorders. However, the computational requirements for the solution of a moderate size gene network amounts to many hours or days of computations. Though there are a number of algorithms suggested by researchers attempting to solve this computationally intensive problem with lower computational efforts, a drastic reduction in computation is yet to achieve. The proposed approach incorporates the modified cuckoo search in S-system model for faster convergence. Means squared error is employed for fitness evaluation. The proposed approach does not require prior biological knowledge about the gene relationships. The performance of the approach is compared with the existing standard approaches such as memetic algorithm based s-system and normal cuckoo search based s-system. The proposed approach is also implemented on real life data set called SOS DNA repair system of Escherichia coli, and found to provide better prediction of relations.