Learning Gene Regulation from Microarray Data via Hidden Markov Models

An important problem in computational biology is prediction of gene regulatory networks. There are many approaches to this problem. However hidden Markov models that are known to show high performance in signal similarity related uses are hard to come by in literature. We have shown through our investigations that this method outperforms correlation method. Furthermore, it is clear that this method can be improved to achieve even higher performance. Hidden Markov models are a reasonable candidate in becoming a useful tool in predicting gene regulatory networks.