Training HMM Structure with Genetic Algorithm for Biological Sequence Analysis
Kyoung-Jae Won, Adam Prügel-Bennett and Anders Krogh
Motivation : Hidden Markov Models (HMMs) are widely used for biological sequence analysis because of their ability to incorporate biological information in their structure. An automatic means of optimising the structure of HMMs would be highly desirable. However, this raises two important issues; firstly, the new HMMs should be biologically interpretable, and secondly we need to control the complexity of the HMM so that it has good generalisation performance on unseen sequences. In this paper we explore the possibility of using a Genetic Algorithm (GA) for optimising the HMM structure. GAs are sufficiently flexible to allow incorporation of other techniques such as Baum-Welch training within their evolutionary cycle. Furthermore, operators which alter the structure of HMMs can be designed to favour interpretable and simple structures.
Results : In this paper a training strategy using Genetic Algorithms is proposed, and it is tested on finding HMM structures for the promoter and coding region of the bacterium C. jejuni. The proposed Genetic Algorithm for Hidden Markov Models (GA-HMM) allows HMMs with different numbers of states to evolve. To prevent over-fitting a separate data set is used for comparing the performance of the HMMs to that used for the Baum-Welch training. The GA-HMM was capable of finding an HMM comparable to a hand-coded HMM designed for the same task, which has previously been published.