MORPH: probabilistic alignment combined with hidden Markov models of cis-regulatory modules.

We first use synthetic data to demonstrate that the new probabilistic model, henceforth called the MORPH framework, can lead to highly significant improvement in (1) alignment accuracy on cis- regulatory sequences, as compared to a state-of-the-art alignment program, and (2) binding site prediction accuracy, as compared to an HMM-based program (Stubb [ 9 ]) that works with a fixed alignment.

also in 132,142 other papers