Prediction Of Protein-coding Genes Research Articles

Exploring the transcription start sites and other genomic features facilitates the accurate identification and annotation of small RNAs across multiple stress conditions in Mycobacterium tuberculosis.

Mycobacterium tuberculosis (MTB) is a pathogen that is known for its ability to persist in harsh environments and cause chronic infections. Understanding the regulatory networks of MTB is crucial for developing effective treatments. Small regulatory RNAs (sRNAs) play important roles in gene expression regulation in all kingdoms of life, and their classification based solely on genomic location can be imprecise due to the computational-based prediction of protein-coding genes in bacteria, which often neglects segments of mRNA such as 5'UTRs, 3'UTRs, and intercistronic regions of operons. To address this issue, our study simultaneously discovered genomic features such as TSSs, UTRs, and operons together with sRNAs in the M. tuberculosis H37Rv strain (ATCC 27294) across multiple stress conditions. Our analysis identified 1,376 sRNA candidates and 8,173 TSSs in MTB, providing valuable insights into its complex regulatory landscape. TSS mapping enabled us to classify these sRNAs into more specific categories, including promoter-associated sRNAs, 5'UTR-derived sRNAs, 3'UTR-derived sRNAs, true intergenic sRNAs, and antisense sRNAs. Three of these sRNA candidates were experimentally validated using 3'-RACE-PCR: predictedRNA_0240, predictedRNA_0325, and predictedRNA_0578. Future characterization and validation are necessary to fully elucidate the functions and roles of these sRNAs in MTB. Our study is the first to simultaneously unravel TSSs and sRNAs in MTB and demonstrate that the identification of other genomic features, such as TSSs, UTRs, and operons, allows for more accurate and specific classification of sRNAs.

The Normal Human Adult Hypothalamus Proteomic Landscape: Rise of Neuroproteomics in Biological Psychiatry and Systems Biology.

The human hypothalamus is central to the regulation of neuroendocrine and neurovegetative systems, as well as modulation of chronobiology and behavioral aspects in human health and disease. Surprisingly, a deep proteomic analysis of the normal human hypothalamic proteome has been missing for such an important organ so far. In this study, we delineated the human hypothalamus proteome using a high-resolution mass spectrometry approach which resulted in the identification of 5349 proteins, while a multiple post-translational modification (PTM) search identified 191 additional proteins, which were missed in the first search. A proteogenomic analysis resulted in the discovery of multiple novel protein-coding regions as we identified proteins from noncoding regions (pseudogenes) and proteins translated from short open reading frames that can be missed using the traditional pipeline of prediction of protein-coding genes as a part of genome annotation. We also identified several PTMs of hypothalamic proteins that may be required for normal hypothalamic functions. Moreover, we observed an enrichment of proteins pertaining to autophagy and adult neurogenesis in the proteome data. We believe that the hypothalamic proteome reported herein would help to decipher the molecular basis for the diverse range of physiological functions attributed to it, as well as its role in neurological and psychiatric diseases. Extensive proteomic profiling of the hypothalamic nuclei would further elaborate on the role and functional characterization of several hypothalamus-specific proteins and pathways to inform future research and clinical discoveries in biological psychiatry, neurology, and system biology.

The parasite Schistocephalus solidus secretes proteins with putative host manipulation functions

BackgroundManipulative parasites are thought to liberate molecules in their external environment, acting as manipulation factors with biological functions implicated in their host’s physiological and behavioural alterations. These manipulation factors are part of a complex mixture called the secretome. While the secretomes of various parasites have been described, there is very little data for a putative manipulative parasite. It is necessary to study the molecular interaction between a manipulative parasite and its host to better understand how such alterations evolve.MethodsHere, we used proteomics to characterize the secretome of a model cestode with a complex life cycle based on trophic transmission. We studied Schistocephalus solidus during the life stage in which behavioural changes take place in its obligatory intermediate fish host, the threespine stickleback (Gasterosteus aculeatus). We produced a novel genome sequence and assembly of S. solidus to improve protein coding gene prediction and annotation for this parasite. We then described the whole worm’s proteome and its secretome during fish host infection using LC–MS/MS.ResultsA total of 2290 proteins were detected in the proteome of S. solidus, and 30 additional proteins were detected specifically in the secretome. We found that the secretome contains proteases, proteins with neural and immune functions, as well as proteins involved in cell communication. We detected receptor-type tyrosine-protein phosphatases, which were reported in other parasitic systems to be manipulation factors. We also detected 12 S. solidus-specific proteins in the secretome that may play important roles in host–parasite interactions.ConclusionsOur results suggest that S. solidus liberates molecules with putative host manipulation functions in the host and that many of them are species-specific.Graphical abstract

A chromosome-level reference genome of the hazelnut, Corylus heterophylla Fisch.

Background Corylus heterophylla Fisch. is a species of the Betulaceae family native to China. As an economically and ecologically important nut tree, C. heterophylla can survive in extremely low temperatures (–30 to –40 °C). To deepen our knowledge of the Betulaceae species and facilitate the use of C. heterophylla for breeding and its genetic improvement, we have sequenced the whole genome of C. heterophylla.FindingsBased on >64.99 Gb (∼175.30×) of Nanopore long reads, we assembled a 370.75-Mb C. heterophylla genome with contig N50 and scaffold N50 sizes of 2.07 and 31.33 Mb, respectively, accounting for 99.23% of the estimated genome size (373.61 Mb). Furthermore, 361.90 Mb contigs were anchored to 11 chromosomes using Hi-C link data, representing 97.61% of the assembled genome sequences. Transcriptomes representing 4 different tissues were sequenced to assist protein-coding gene prediction. A total of 27,591 protein-coding genes were identified, of which 92.02% (25,389) were functionally annotated. The phylogenetic analysis showed that C. heterophylla is close to Ostrya japonica, and they diverged from their common ancestor ∼52.79 million years ago.ConclusionsWe generated a high-quality chromosome-level genome of C. heterophylla. This genome resource will promote research on the molecular mechanisms of how the hazelnut responds to environmental stresses and serves as an important resource for genome-assisted improvement in cold and drought resistance of the Corylus genus.

Understanding the causes of errors in eukaryotic protein-coding gene prediction: a case study of primate proteomes

BackgroundRecent advances in sequencing technologies have led to an explosion in the number of genomes available, but accurate genome annotation remains a major challenge. The prediction of protein-coding genes in eukaryotic genomes is especially problematic, due to their complex exon–intron structures. Even the best eukaryotic gene prediction algorithms can make serious errors that will significantly affect subsequent analyses.ResultsWe first investigated the prevalence of gene prediction errors in a large set of 176,478 proteins from ten primate proteomes available in public databases. Using the well-studied human proteins as a reference, a total of 82,305 potential errors were detected, including 44,001 deletions, 27,289 insertions and 11,015 mismatched segments where part of the correct protein sequence is replaced with an alternative erroneous sequence. We then focused on the mismatched sequence errors that cause particular problems for downstream applications. A detailed characterization allowed us to identify the potential causes for the gene misprediction in approximately half (5446) of these cases. As a proof-of-concept, we also developed a simple method which allowed us to propose improved sequences for 603 primate proteins.ConclusionsGene prediction errors in primate proteomes affect up to 50% of the sequences. Major causes of errors include undetermined genome regions, genome sequencing or assembly issues, and limitations in the models used to represent gene exon–intron structures. Nevertheless, existing genome sequences can still be exploited to improve protein sequence quality. Perspectives of the work include the characterization of other types of gene prediction errors, as well as the development of a more comprehensive algorithm for protein sequence error correction.

UniRule: a unified rule resource for automatic annotation in the UniProt Knowledgebase.

MotivationThe number of protein records in the UniProt Knowledgebase (UniProtKB: https://www.uniprot.org) continues to grow rapidly as a result of genome sequencing and the prediction of protein-coding genes. Providing functional annotation for these proteins presents a significant and continuing challenge.ResultsIn response to this challenge, UniProt has developed a method of annotation, known as UniRule, based on expertly curated rules, which integrates related systems (RuleBase, HAMAP, PIRSR, PIRNR) developed by the members of the UniProt consortium. UniRule uses protein family signatures from InterPro, combined with taxonomic and other constraints, to select sets of reviewed proteins which have common functional properties supported by experimental evidence. This annotation is propagated to unreviewed records in UniProtKB that meet the same selection criteria, most of which do not have (and are never likely to have) experimentally verified functional annotation. Release 2020_01 of UniProtKB contains 6496 UniRule rules which provide annotation for 53 million proteins, accounting for 30% of the 178 million records in UniProtKB. UniRule provides scalable enrichment of annotation in UniProtKB.Availability and implementationUniRule rules are integrated into UniProtKB and can be viewed at https://www.uniprot.org/unirule/. UniRule rules and the code required to run the rules, are publicly available for researchers who wish to annotate their own sequences. The implementation used to run the rules is known as UniFIRE and is available at https://gitlab.ebi.ac.uk/uniprot-public/unifire.

De Novo Transcriptome Assembly of Cucurbita Pepo L. Leaf Tissue Infested by Aphis Gossypii

Zucchini (Cucurbita pepo L.), extensively cultivated in temperate areas, belongs to the Cucurbitaceae family and it is a species with great economic value. One major threat related to zucchini cultivation is the damage imposed by the cotton/melon aphid Aphis gossypii Glover (Homoptera: Aphididae). We performed RNA-sequencing on cultivar “San Pasquale” leaves, uninfested and infested by A. gossypii, that were collected at three time points (24, 48, and 96 h post infestation). Then, we combined all high-quality reads for de novo assembly of the transcriptome. This resource was primarily established to be used as a reference for gene expression studies in order to investigate the transcriptome reprogramming of zucchini plants following aphid infestation. In addition, raw reads will be valuable for new experiments based on the latest bioinformatic tools and analytical approaches. The assembled transcripts will serve as an important reference for sequence-based studies and for primer design. Both datasets can be used to support/improve the prediction of protein-coding genes in the zucchini genome, which has been recently released into the public domain.

Gapless genome assembly of Colletotrichum higginsianum reveals chromosome structure and association of transposable elements with secondary metabolite gene clusters

BackgroundThe ascomycete fungus Colletotrichum higginsianum causes anthracnose disease of brassica crops and the model plant Arabidopsis thaliana. Previous versions of the genome sequence were highly fragmented, causing errors in the prediction of protein-coding genes and preventing the analysis of repetitive sequences and genome architecture.ResultsHere, we re-sequenced the genome using single-molecule real-time (SMRT) sequencing technology and, in combination with optical map data, this provided a gapless assembly of all twelve chromosomes except for the ribosomal DNA repeat cluster on chromosome 7. The more accurate gene annotation made possible by this new assembly revealed a large repertoire of secondary metabolism (SM) key genes (89) and putative biosynthetic pathways (77 SM gene clusters). The two mini-chromosomes differed from the ten core chromosomes in being repeat- and AT-rich and gene-poor but were significantly enriched with genes encoding putative secreted effector proteins. Transposable elements (TEs) were found to occupy 7% of the genome by length. Certain TE families showed a statistically significant association with effector genes and SM cluster genes and were transcriptionally active at particular stages of fungal development. All 24 subtelomeres were found to contain one of three highly-conserved repeat elements which, by providing sites for homologous recombination, were probably instrumental in four segmental duplications.ConclusionThe gapless genome of C. higginsianum provides access to repeat-rich regions that were previously poorly assembled, notably the mini-chromosomes and subtelomeres, and allowed prediction of the complete SM gene repertoire. It also provides insights into the potential role of TEs in gene and genome evolution and host adaptation in this asexual pathogen.

Well-characterized sequence features of eukaryote genomes and implications for ab initio gene prediction

In silico analysis of DNA sequences is an important area of computational biology in the post-genomic era. Over the past two decades, computational approaches for ab initio prediction of gene structure from genome sequence alone have largely facilitated our understanding on a variety of biological questions. Although the computational prediction of protein-coding genes has already been well-established, we are also facing challenges to robustly find the non-coding RNA genes, such as miRNA and lncRNA. Two main aspects of ab initio gene prediction include the computed values for describing sequence features and used algorithm for training the discriminant function, and by which different combinations are employed into various bioinformatic tools. Herein, we briefly review these well-characterized sequence features in eukaryote genomes and applications to ab initio gene prediction. The main purpose of this article is to provide an overview to beginners who aim to develop the related bioinformatic tools.

Distinctive expansion of gene families associated with plant cell wall degradation, secondary metabolism, and nutrient uptake in the genomes of grapevine trunk pathogens.

BackgroundTrunk diseases threaten the longevity and productivity of grapevines in all viticulture production systems. They are caused by distantly-related fungi that form chronic wood infections. Variation in wood-decay abilities and production of phytotoxic compounds are thought to contribute to their unique disease symptoms. We recently released the draft sequences of Eutypa lata, Neofusicoccum parvum and Togninia minima, causal agents of Eutypa dieback, Botryosphaeria dieback and Esca, respectively. In this work, we first expanded genomic resources to three important trunk pathogens, Diaporthe ampelina, Diplodia seriata, and Phaeomoniella chlamydospora, causal agents of Phomopsis dieback, Botryosphaeria dieback, and Esca, respectively. Then we integrated all currently-available information into a genome-wide comparative study to identify gene families potentially associated with host colonization and disease development.ResultsThe integration of RNA-seq, comparative and ab initio approaches improved the protein-coding gene prediction in T. minima, whereas shotgun sequencing yielded nearly complete genome drafts of Dia. ampelina, Dip. seriata, and P. chlamydospora. The predicted proteomes of all sequenced trunk pathogens were annotated with a focus on functions likely associated with pathogenesis and virulence, namely (i) wood degradation, (ii) nutrient uptake, and (iii) toxin production. Specific patterns of gene family expansion were described using Computational Analysis of gene Family Evolution, which revealed lineage-specific evolution of distinct mechanisms of virulence, such as specific cell wall oxidative functions and secondary metabolic pathways in N. parvum, Dia. ampelina, and E. lata. Phylogenetically-informed principal component analysis revealed more similar repertoires of expanded functions among species that cause similar symptoms, which in some cases did not reflect phylogenetic relationships, thereby suggesting patterns of convergent evolution.ConclusionsThis study describes the repertoires of putative virulence functions in the genomes of ubiquitous grapevine trunk pathogens. Gene families with significantly faster rates of gene gain can now provide a basis for further studies of in planta gene expression, diversity by genome re-sequencing, and targeted reverse genetic approaches. The functional validation of potential virulence factors will lead to a more comprehensive understanding of the mechanisms of pathogenesis and virulence, which ultimately will enable the development of accurate diagnostic tools and effective disease management.Electronic supplementary materialThe online version of this article (doi:10.1186/s12864-015-1624-z) contains supplementary material, which is available to authorized users.

개 회충 게놈 응용 사례에서 공개용 분석 툴을 사용한 드래프트 게놈 어셈블리 생성

It has become possible for small scale laboratories to interpret large scale genomic DNA, thanks to the reduction of the sequencing cost by the development of next generation sequencing (NGS). De novo assembly is a method which creates a putative original sequence by reconstructing reads without using a reference sequence. There have been various study results on de novo assembly, however, it is still difficult to get the desired results even by using the same assembly procedures and the analysis tools which were suggested in the studies reported. This is mainly because there are no specific guidelines for the assembly procedures or know-hows for the use of such analysis tools. In this study, to resolve these problems, we introduce steps to finding whole genome of an unknown DNA via NGS technology and de novo assembly, while providing the pros and cons of the various analysis tools used in each step. We used 350Mbp of Toxocara canis DNA as an application case for the detailed explanations of each stated step. We also extend our works for prediction of protein-coding genes and their functions from the draft genome sequence by comparing its homology with reference sequences of other nematodes.

SIBIS: a Bayesian model for inconsistent protein sequence estimation

The prediction of protein coding genes is a major challenge that depends on the quality of genome sequencing, the accuracy of the model used to elucidate the exonic structure of the genes and the complexity of the gene splicing process leading to different protein variants. As a consequence, today's protein databases contain a huge amount of inconsistency, due to both natural variants and sequence prediction errors. We have developed a new method, called SIBIS, to detect such inconsistencies based on the evolutionary information in multiple sequence alignments. A Bayesian framework, combined with Dirichlet mixture models, is used to estimate the probability of observing specific amino acids and to detect inconsistent or erroneous sequence segments. We evaluated the performance of SIBIS on a reference set of protein sequences with experimentally validated errors and showed that the sensitivity is significantly higher than previous methods, with only a small loss of specificity. We also assessed a large set of human sequences from the UniProt database and found evidence of inconsistency in 48% of the previously uncharacterized sequences. We conclude that the integration of quality control methods like SIBIS in automatic analysis pipelines will be critical for the robust inference of structural, functional and phylogenetic information from these sequences. Source code, implemented in C on a linux system, and the datasets of protein sequences are freely available for download at http://www.lbgi.fr/∼julie/SIBIS.

Improved structural annotation of protein-coding genes in the Meloidogyne hapla genome using RNA-Seq.

As high-throughput cDNA sequencing (RNA-Seq) is increasingly applied to hypothesis-driven biological studies, the prediction of protein coding genes based on these data are usurping strictly in silico approaches. Compared with computationally derived gene predictions, structural annotation is more accurate when based on biological evidence, particularly RNA-Seq data. Here, we refine the current genome annotation for the Meloidogyne hapla genome utilizing RNA-Seq data. Published structural annotation defines 14 420 protein-coding genes in the M. hapla genome. Of these, 25% (3751) were found to exhibit some incongruence with RNA-Seq data. Manual annotation enabled these discrepancies to be resolved. Our analysis revealed 544 new gene models that were missing from the prior annotation. Additionally, 1457 transcribed regions were newly identified on the ends of as-yet-unjoined contigs. We also searched for trans-spliced leaders, and based on RNA-Seq data, identified genes that appear to be trans-spliced. Four 22-bp trans-spliced leaders were identified using our pipeline, including the known trans-spliced leader, which is the M. hapla ortholog of SL1. In silico predictions of trans-splicing were validated by comparison with earlier results derived from an independent cDNA library constructed to capture trans-spliced transcripts. The new annotation, which we term HapPep5, is publically available at www.hapla.org.

WebAUGUSTUS--a web service for training AUGUSTUS and predicting genes in eukaryotes

The prediction of protein coding genes is an important step in the annotation of newly sequenced and assembled genomes. AUGUSTUS is one of the most accurate tools for eukaryotic gene prediction. Here, we present WebAUGUSTUS, a web interface for training AUGUSTUS and predicting genes with AUGUSTUS. Depending on the needs of the user, WebAUGUSTUS generates training gene structures automatically. Besides a genome file, either a file with expressed sequence tags or a file with protein sequences is required for this step. Alternatively, it is possible to submit an externally generated training gene structure file and a genome file. The web service optimizes AUGUSTUS parameters and predicts genes with those parameters. WebAUGUSTUS is available at http://bioinf.uni-greifswald.de/webaugustus.

Re-Annotation of Protein-Coding Genes in 10 Complete Genomes of Neisseriaceae Family by Combining Similarity-Based and Composition-Based Methods

In this paper, we performed a comprehensive re-annotation of protein-coding genes by a systematic method combining composition- and similarity-based approaches in 10 complete bacterial genomes of the family Neisseriaceae. First, 418 hypothetical genes were predicted as non-coding using the composition-based method and 413 were eliminated from the gene list. Both the scatter plot and cluster of orthologous groups (COG) fraction analyses supported the result. Second, from 20 to 400 hypothetical proteins were assigned with functions in each of the 10 strains based on the homology search. Among newly assigned functions, 397 are so detailed to have definite gene names. Third, 106 genes missed by the original annotations were picked up by an ab initio gene finder combined with similarity alignment. Transcriptional experiments validated the effectiveness of this method in Laribacter hongkongensis and Chromobacterium violaceum. Among the 106 newly found genes, some deserve particular interests. For example, 27 transposases were newly found in Neiserria meningitidis alpha14. In Neiserria gonorrhoeae NCCP11945, four new genes with putative functions and definite names (nusG, rpsN, rpmD and infA) were found and homologues of them usually are essential for survival in bacteria. The updated annotations for the 10 Neisseriaceae genomes provide a more accurate prediction of protein-coding genes and a more detailed functional information of hypothetical proteins. It will benefit research into the lifestyle, metabolism, environmental adaption and pathogenicity of the Neisseriaceae species. The re-annotation procedure could be used directly, or after the adaption of detailed methods, for checking annotations of any other bacterial or archaeal genomes.

Genomic law guided gene prediction in fungi and metazoans

Protein coding gene prediction by computational approaches is a fundamental step for genome annotation. However, it is a challenge to accurately predict eukaryotic genes in silico. By surveying the model genomes, we found that the Spearman's rank correlation coefficient between the number of experimental-verified genes and the size of genomes was 0.96 for all eukaryotes except plants, indicating the relationship between genome size and the number of coding genes can be expressed with a monotonic function. Regression analysis found that the relationship of total protein coding genes over genome size followed a logarithmic equation. We integrated the equation into ab initio gene prediction software to guide the gene prediction by constraining the total number of predicted genes. We evaluated the software in three eukaryotic genomes. Results showed that >90% of false positive predictions were removed while >80% of true positives were retained, resulting in much higher specificity.

A Modified Statistically Optimal Null Filter Method for Recognizing Protein-coding Regions

Computer-aided protein-coding gene prediction in uncharacterized genomic DNA sequences is one of the most important issues of biological signal processing. A modified filter method based on a statistically optimal null filter (SONF) theory is proposed for recognizing protein-coding regions. The square deviation gain (SDG) between the input and output of the model is used to identify the coding regions. The effective SDG amplification model with Class I and Class II enhancement is designed to suppress the non-coding regions. Also, an evaluation algorithm has been used to compare the modified model with most gene prediction methods currently available in terms of sensitivity, specificity and precision. The performance for identification of protein-coding regions has been evaluated at the nucleotide level using benchmark datasets and 91.4%, 96%, 93.7% were obtained for sensitivity, specificity and precision, respectively. These results suggest that the proposed model is potentially useful in gene finding field, which can help recognize protein-coding regions with higher precision and speed than present algorithms.

RNA-Seq improves annotation of protein-coding genes in the cucumber genome

BackgroundAs more and more genomes are sequenced, genome annotation becomes increasingly important in bridging the gap between sequence and biology. Gene prediction, which is at the center of genome annotation, usually integrates various resources to compute consensus gene structures. However, many newly sequenced genomes have limited resources for gene predictions. In an effort to create high-quality gene models of the cucumber genome (Cucumis sativus var. sativus), based on the EVidenceModeler gene prediction pipeline, we incorporated the massively parallel complementary DNA sequencing (RNA-Seq) reads of 10 cucumber tissues into EVidenceModeler. We applied the new pipeline to the reassembled cucumber genome and included a comparison between our predicted protein-coding gene sets and a published set.ResultsThe reassembled cucumber genome, annotated with RNA-Seq reads from 10 tissues, has 23, 248 identified protein-coding genes. Compared with the published prediction in 2009, approximately 8, 700 genes reveal structural modifications and 5, 285 genes only appear in the reassembled cucumber genome. All the related results, including genome sequence and annotations, are available at http://cmb.bnu.edu.cn/Cucumis_sativus_v20/.ConclusionsWe conclude that RNA-Seq greatly improves the accuracy of prediction of protein-coding genes in the reassembled cucumber genome. The comparison between the two gene sets also suggests that it is feasible to use RNA-Seq reads to annotate newly sequenced or less-studied genomes.

Relationship of oxidative stress and endothelial dysfunction in sleep apnoea

The aim of the present study was to evaluate ischaemic reactive hyperaemia (IRH) in obstructive sleep apnoea (OSA) and its relationship with oxidative stress. We studied 69 consecutive patients referred to our Sleep Unit (Reina Sofia University Hospital, Cordoba, Spain). Patients with chronic diseases or those taking medication were excluded. IRH was assessed before and after polysomnography. Morning IRH and oxidative stress markers were compared between patients with (apnoea-hypopnoea index (AHI) ≥ 5) and without (AHI < 5) OSA. Measurements were repeated in 25 severe OSA patients after continuous positive airway pressure (CPAP) therapy. We included 46 OSA patients (mean ± sd AHI 49 ± 32.1) and 23 non-OSA subjects (AHI 3 ± 0.9). The OSA patients showed a significant worsening of morning IRH, and a significant increase in malondialdehyde and 8-hydroxydeoxyguanosine levels. Only the oxygen desaturation index independently explained morning IRH, while malondialdehyde levels showed a weak effect on IRH. In severe OSA patients, IRH improved significantly after CPAP treatment, as did malondialdehyde, 8-hydroxydeoxyguanosine and protein carbonyl levels. In OSA patients, endothelial dysfunction and oxidative stress were observed, and IRH worsened after sleep. The increase in oxidative stress was not associated with IRH, while intermittent hypoxia was strongly associated with IRH. In severe OSA patients, CPAP treatment improved oxidative stress and endothelial function.

MGene.web: a web service for accurate computational gene finding

We describe mGene.web, a web service for the genome-wide prediction of protein coding genes from eukaryotic DNA sequences. It offers pre-trained models for the recognition of gene structures including untranslated regions in an increasing number of organisms. With mGene.web, users have the additional possibility to train the system with their own data for other organisms on the push of a button, a functionality that will greatly accelerate the annotation of newly sequenced genomes. The system is built in a highly modular way, such that individual components of the framework, like the promoter prediction tool or the splice site predictor, can be used autonomously. The underlying gene finding system mGene is based on discriminative machine learning techniques and its high accuracy has been demonstrated in an international competition on nematode genomes. mGene.web is available at http://www.mgene.org/web, it is free of charge and can be used for eukaryotic genomes of small to moderate size (several hundred Mbp).