Abstract
We have made several steps towards creating fast and accurate algorithm for gene prediction in eukaryotic genomes. First, we introduced an automated method for efficient ab initio gene finding, GeneMark-ES, with parameters trained in iterative unsupervised mode. Next, in GeneMark-ET we proposed a method of integration of unsupervised training with information on intron positions revealed by mapping short RNA reads. Now we describe GeneMark-EP, a tool that utilizes another source of external information, a protein database, readily available prior to a start of a sequencing project. The new algorithm and software tool integrate information produced by proteins spliced aligned to genomic regions into model training and gene prediction steps. A specialized pipeline, ProtHint, makes processing the results of mapping of multiple proteins to a genomic region where a protein from the same family is likely encoded. GeneMark-EP uses the hints from ProtHint to improve estimation of model parameters as well as to adjust co-ordinates of predicted genes if they disagree with the most reliable hints (the -EP+ mode). Tests conducted with GeneMark-EP and -EP+ have demonstrated that the gene prediction accuracy is higher than one of GeneMark-ES, particularly in large eukaryotic genomes.
Footnotes
↵† Joint first authors
