AnABlast: a newin silicostrategy for the genome-wide search of novel genes and fossil regions

Abstract

Genome annotation, assisted by computer programs, is one of the great advances in modern biology. Nevertheless, the in silico identification of small and complex coding sequences is still challenging. We observed that amino acid sequences inferred from coding—but rarely from non-coding—DNA sequences accumulated alignments in low-stringency BLAST searches, suggesting that this alignments accumulation could be used to highlight coding regions in sequenced DNA. To investigate this possibility, we developed a computer program (AnABlast) that generates profiles of accumulated alignments in query amino acid sequences using a low-stringency BLAST strategy. To validate this approach, all six-frame translations of DNA sequences between every two annotated exons of the fission yeast genome were analysed with AnABlast. AnABlast-generated profiles identified three new copies of known genes, and four new genes supported by experimental evidence. New pseudogenes, ancestral carboxyl- and amino-terminal subtractions, complex gene rearrangements, and ancient fragments of mitDNA and of bacterial origin, were also inferred. Thus, this novel in silico approach provides a powerful tool to uncover new genes, as well as fossil-coding sequences, thus providing insight into the evolutionary history of annotated genomes.

Ancient evolutionary signals of protein-coding sequences allow the discovery of new genes in the Drosophila melanogaster genome

“…This observation agree with previous evidences in mammals suggesting a higher frequency of evolutionary fixation for deletion than for insertion mutations [18,19], a trend that have also been found in the D. melanogaster genome [20]. Remarkably, AnABlast coding-signal are sometimes found in the ends of well-known genes, overlapping with the right reading frame and suggesting than C-terminus and N-terminus of genes are subjected to evolutionary contractions and expansions that are efficiently identified by AnABlast [10].…”

“…It leads to the necessary development of new algorithms based on different ideas [12,13]. In this context, we proposed a new in silico strategy, named AnABlast that uses low-score alignments coming from multiple non-redundant proteins [10]. As shown in this study, in agreement with previous reports [14][15][16], these Green color represents protomotif accumulation in the forward strand, and red color in the reverse strand.…”

“…To optimize AnABlast parameters for the identification of new exons and genes, the distribution of both true and false positive coding-signals were evaluated at different peak-height thresholds. AnABlast profiles depend on the bit-score value used to restrict alignment significance during the BLAST search, therefore, in addition to the value of bit-score 30, previously used by AnABlast [10], the evaluation was carried out also using the more and less restrictive bit-scores of 29 and 31 respectively. Regardless of the taken score, true positive coding-signals account for the highest peak-height ( Fig.…”

“…Traditional gene finders are routinely based on both significant sequence similarity and sequence signatures such as those used to define open reading frames (ORF), signals involved in splicing [8], or combined protocols to get better results [9]. Among the new proposed methods, we have previously shown that accumulation of low-score alignments, which would represent footprints of ancient sequences, highlights present and ancient protein-coding regions which are hard to discover by conventional methods [10]. Briefly, this novel computational approach, that we named AnA-Blast, compares the putative amino acid sequences from the six reading frames of a genomic sequence against a nonredundant protein database, and collects the matches, including low-score alignments, which we call protomotifs.…”

“…We previously showed that this strategy is useful to discover putative protein-coding regions. It was tested with both intergenic and intronic sequences from the fission yeast Schizosaccharomyces pombe, and 18 putative genes were predicted [10]. But even though these predictions had computational support and some translation evidences, prediction parameters were empirically established by training the algorithm with S. pombe curated genes.…”