Using intron position conservation for homology-based gene prediction

Abstract

Annotation of protein-coding genes is very important in bioinformatics and biology and has a decisive influence on many downstream analyses. Homology-based gene prediction programs allow for transferring knowledge about protein-coding genes from an annotated organism to an organism of interest.Here, we present a homology-based gene prediction program called GeMoMa. GeMoMa utilizes the conservation of intron positions within genes to predict related genes in other organisms. We assess the performance of GeMoMa and compare it with state-of-the-art competitors on plant and animal genomes using an extended best reciprocal hit approach. We find that GeMoMa often makes more precise predictions than its competitors yielding a substantially increased number of correct transcripts. Subsequently, we exemplarily validate GeMoMa predictions using Sanger sequencing. Finally, we use RNA-seq data to compare the predictions of homology-based gene prediction programs, and find again that GeMoMa performs well.Hence, we conclude that exploiting intron position conservation improves homology-based gene prediction, and we make GeMoMa freely available as command-line tool and Galaxy integration.

The pomegranate (Punica granatum L.) draft genome dissects genetic divergence between soft‐ and hard‐seeded cultivars

“…Open reading frames within the transcripts were predicted with the TransDecoder program (version 2.0). Arabidopsis_thaliana, Sorghum_bicolor , Malus × domestica , Vitis_vinifera and Eucalyptus grandis were used for the homology‐based predictions with GeneMoMa (Keilwagen et al ., ). Potential candidate gene sets generated by PASA (version 2.0.2) (Haas et al ., ) based on cDNA sequences as well as homology‐based gene models were used for analyses with de novo prediction software packages (i.e.…”

Comparison ofArachis monticolawith Diploid and Cultivated Tetraploid Genomes Reveals Asymmetric Subgenome Evolution and Improvement of Peanut

“…Three pieces of software, including Genscan (http://genes.mit.edu/GENSCAN.html), Augustus v2.4, and GlimmerHMM v3.0.4, were used for de novo prediction. The homolog‐based prediction was refined by GeneWise v2.4.1 and GeMoMa v1.3.1 . Transcriptome data that were generated from pooled tissues of leaf, stem, and root of wild peanut were mapped and assembled using Hisat v2.0.4 and Stringtie v1.2.3, respectively.…”

“…The homolog-based prediction was refined by GeneWise v2.4.1 and GeMoMa v1.3.1. [76,77] Transcriptome data that were generated from pooled tissues of leaf, stem, and root of wild peanut were mapped and assembled using Hisat v2.0.4 [78] and Stringtie v1.2.3, [79] respectively. Unigenes were aligned to the genome assembly using BLAT [80] and then filtered using PASA2.0.4.…”

Functional insights from the GC-poor genomes of two aphid parasitoids,Aphidius erviandLysiphlebus fabarum

“…GeMoMa (Keilwagen et al 2016 melanogaster protein sequence was used in blast similarity searches against the two 1063 predicted wasp proteomes. The best match was retained, and its protein sequence was 1064 used to perform a new blast search using the NCBI non-redundant protein sequence 1065 database to confirm the similarity with the D. melanogaster sequence.…”