New Eukaryotic Gene Finder Integrating RNA-Seq and Protein Mapping into ab initio Algorithm with Semi-Supervised Training

We present a new algorithm, which i/ utilizes evidence derived from RNA-Seq and protein mapping to assist self-training of ab initio gene prediction algorithm as well as ii/ uses this external evidence to generate consistent prediction across the whole genome.

At the training stage the algorithm performs iterative estimation of parameters of hidden semi-Markov model. RNA-Seq reads mapping of splice junctions is notoriously noisy (low specificity) while protein mapping of closely related protein sequences is patchy (low sensitivity). Still it is possible to combine both types of mapping with self-training procedure in an optimal way:  to filter out RNA-Seq derived false positive introns while filling in gene structure elements missed by protein mapping. At the last step of iterations, the final gene prediction step, high scored RNA-Seq read and protein mappings are integrated into the final prediction.

Algorithm is robust with respect to assembly fragmentation and genome size. It can accommodate any amount, small or large, of RNA-Seq read and protein mapping information into training and gene prediction. The algorithm is a new addition to the GeneMark line of gene finders ( topaz.gatech.edu/GeneMark ).