A Method to Improve the Performance of Translation Start Site Detection and Its Application for Gene Finding

摘要

Correct signal identification is an important step for all ab initio gene finders. The aim of this work is to introduce a new method for detecting translation start sites in genomic DNA sequences. By using interpolated context Markov model to capture the coding potential of the region immediately following a putative start codon, the novel method described achieves an 84% accuracy of the start codon recognition. The implementation of this technique into GlimmerM succeeds in improving the sensitivity of the gene prediction by 5%.