%0 Journal Article
%T An unsupervised classification scheme for improving predictions of prokaryotic TIS
%A Maike Tech
%A Peter Meinicke
%J BMC Bioinformatics
%D 2006
%I BioMed Central
%R 10.1186/1471-2105-7-121
%X We introduce a clustering algorithm for completely unsupervised scoring of potential TIS, based on positionally smoothed probability matrices. The algorithm requires an initial gene prediction and the genomic sequence of the organism to perform the reannotation. As compared with other methods for improving predictions of gene starts in bacterial genomes, our approach is not based on any specific assumptions about prokaryotic TIS. Despite the generality of the underlying algorithm, the prediction rate of our method is competitive on experimentally verified test data from E. coli and B. subtilis. Regarding genomes with high G+C content, in contrast to some previously proposed methods, our algorithm also provides good performance on P. aeruginosa, B. pseudomallei and R. solanacearum.On reliable test data we showed that our method provides good results in post-processing the predictions of the widely-used program GLIMMER. The underlying clustering algorithm is robust with respect to variations in the initial TIS annotation and does not require specific assumptions about prokaryotic gene starts. These features are particularly useful on genomes with high G+C content. The algorithm has been implemented in the tool £¿TICO£¿(TIs COrrector) which is publicly available from our web site.Recent publications have shown that gene prediction in prokaryotes is still a challenging problem in bioinformatics. While existing gene finders [1-3] are likely to identify coding regions associated with open reading frames (ORF) of statistically significant length, the prediction of the true translation initiation sites (TIS) is insufficient in many cases [4-6]. In particular, for genomes with a high G+C content the prediction of TIS has been shown to be of low quality [6].In order to cope with this insufficiency of conventional gene finders, several methods have been proposed for improving the predictions of prokaryotic TIS. Common approaches require prior knowledge about the characteristics
%U http://www.biomedcentral.com/1471-2105/7/121