Search

Recent technological advances have opened unprecedented opportunities for large-scale sequencing and analysis of populations of pathogenic species in disease outbreaks, as well as for large-scale diversity studies aimed at expanding our knowledge across the whole domain of prokaryotes. To meet the challenge of timely interpretation of structure, function and meaning of this vast genetic information, a comprehensive approach to automatic genome annotation is critically needed. In collaboration with Georgia Tech, NCBI has developed a new approach to genome annotation that combines alignment based methods with methods of predicting protein-coding and RNA genes and other functional elements directly from sequence. A new gene finding tool, GeneMarkS+, uses the combined evidence of protein and RNA placement by homology as an initial map of annotation to generate and modify ab initio gene predictions across the whole genome. Thus, the new NCBI's Prokaryotic Genome Annotation Pipeline (PGAP) relies more on sequence similarity when confident comparative data are available, while it relies more on statistical predictions in the absence of external evidence. The pipeline provides a framework for generation and analysis of annotation on the full breadth of prokaryotic taxonomy. For additional information on PGAP see https://www.ncbi.nlm.nih.gov/genome/annotation_prok/ and the NCBI Handbook, https://www.ncbi.nlm.nih.gov/books/NBK174280/.

Finding new protein-coding genes is one of the most important goals of eukaryotic genome sequencing projects. However, genomic organization of novel eukaryotic genomes is diverse and ab initio gene finding tools tuned up for previously studied species are rarely suitable for efficacious gene hunting in DNA sequences of a new genome. Gene identification methods based on cDNA and expressed sequence tag (EST) mapping to genomic DNA or those using alignments to closely related genomes rely either on existence of abundant cDNA and EST data and/or availability on reference genomes. Conventional statistical ab initio methods require large training sets of validated genes for estimating gene model parameters. In practice, neither one of these types of data may be available in sufficient amount until rather late stages of the novel genome sequencing. Nevertheless, we have shown that gene finding in eukaryotic genomes could be carried out in parallel with statistical models estimation directly from yet anonymous genomic DNA. The suggested method of parallelization of gene prediction with the model parameters estimation follows the path of the iterative Viterbi training. Rounds of genomic sequence labeling into coding and non-coding regions are followed by the rounds of model parameters estimation. Several dynamically changing restrictions on the possible range of model parameters are added to filter out fluctuations in the initial steps of the algorithm that could redirect the iteration process away from the biologically relevant point in parameter space. Tests on well-studied eukaryotic genomes have shown that the new method performs comparably or better than conventional methods where the supervised model training precedes the gene prediction step. Several novel genomes have been analyzed and biologically interesting findings are discussed. Thus, a self-training algorithm that had been assumed feasible only for prokaryotic genomes has now been developed for ab initio eukaryotic gene identification.

The task of gene identification frequently confronting researchers working with both novel and well studied genomes can be conveniently and reliably solved with the help of the GeneMark web software (http://opal.biology.gatech.edu/GeneMark/). The website provides interfaces to the GeneMark family of programs designed and tuned for gene prediction in prokaryotic, eukaryotic and viral genomic sequences. Currently, the server allows the analysis of nearly 200 prokaryotic and >10 eukaryotic genomes using species-specific versions of the software and pre-computed gene models. In addition, genes in prokaryotic sequences from novel genomes can be identified using models derived on the spot upon sequence submission, either by a relatively simple heuristic approach or by the full-fledged self-training program GeneMarkS. A database of reannotations of >1000 viral genomes by the GeneMarkS program is also available from the web site. The GeneMark website is frequently updated to provide the latest versions of the software and gene models.

We further investigated the statistical features of the three classes of Escherichia coli genes that have been previously delineated by factorial correspondence analysis and dynamic clustering methods. A phased Markov model for a nucleotide sequence of each gene class was developed and employed for gene prediction using the GeneMark program. The protein-coding region prediction accuracy was determined for class-specific Markov models of different orders when the programs implementing these models were applied to gene sequences from the same or other classes. It is shown that at least two training sets and two program versions derived for different classes of E. coli genes are necessary in order to achieve a high accuracy of coding region prediction for uncharacterized sequences. Some annotated E. coli genes from Class I and Class III are shown to be spurious, whereas many open reading frames (ORFs) that have not been annotated in GenBank as genes are predicted to encode proteins. The amino acid sequences of the putative products of these ORFs initially did not show similarity to already known proteins. However, conserved regions have been identified in several of them by screening the latest entries in protein sequence databases and applying methods for motif search, while some other of these new genes have been identified in independent experiments.

The unannotated regions of the Escherichia coli genome DNA sequence from the EcoSeq6 database, totaling 1,278 'intergenic' sequences of the combined length of 359,279 basepairs, were analyzed using computer-assisted methods with the aim of identifying putative unknown genes. The proposed strategy for finding new genes includes two key elements: i) prediction of expressed open reading frames (ORFs) using the GeneMark method based on Markov chain models for coding and non-coding regions of Escherichia coli DNA, and ii) search for protein sequence similarities using programs based on the BLAST algorithm and programs for motif identification. A total of 354 putative expressed ORFs were predicted by GeneMark. Using the BLASTX and TBLASTN programs, it was shown that 208 ORFs located in the unannotated regions of the E. coli chromosome are significantly similar to other protein sequences. Identification of 182 ORFs as probable genes was supported by GeneMark and BLAST, comprising 51.4% of the GeneMark 'hits' and 87.5% of the BLAST 'hits'. 73 putative new genes, comprising 20.6% of the GeneMark predictions, belong to ancient conserved protein families that include both eubacterial and eukaryotic members. This value is close to the overall proportion of highly conserved sequences among eubacterial proteins, indicating that the majority of the putative expressed ORFs that are predicted by GeneMark, but have no significant BLAST hits, nevertheless are likely to be real genes. The majority of the putative genes identified by BLAST search have been described since the release of the EcoSeq6 database, but about 70 genes have not been detected so far. Among these new identifications are genes encoding proteins with a variety of predicted functions including dehydrogenases, kinases, several other metabolic enzymes, ATPases, rRNA methyltransferases, membrane proteins, and different types of regulatory proteins.

We describe an algorithm for gene identification in DNA sequences derived from shotgun sequencing of microbial communities. Accurate ab initio gene prediction in a short nucleotide sequence of anonymous origin is hampered by uncertainty in model parameters. While several machine learning approaches could be proposed to bypass this difficulty, one effective method is to estimate parameters from dependencies, formed in evolution, between frequencies of oligonucleotides in protein-coding regions and genome nucleotide composition. Original version of the method was proposed in 1999 and has been used since for (i) reconstructing codon frequency vector needed for gene finding in viral genomes and (ii) initializing parameters of self-training gene finding algorithms. With advent of new prokaryotic genomes en masse it became possible to enhance the original approach by using direct polynomial and logistic approximations of oligonucleotide frequencies, as well as by separating models for bacteria and archaea. These advances have increased the accuracy of model reconstruction and, subsequently, gene prediction. We describe the refined method and assess its accuracy on known prokaryotic genomes split into short sequences. Also, we show that as a result of application of the new method, several thousands of new genes could be added to existing annotations of several human and mouse gut metagenomes.

Massive parallel sequencing of RNA transcripts by the next generation technology (RNA-Seq) is a powerful method of generating critically important data for discovery of structure and function of eukaryotic genes. The transcripts may or may not carry protein-coding regions. If protein coding region is present, it should be a continuous (spliced) open reading frame. Gene finding in transcripts can be done by statistical (alignment-free) as well as by alignment based methods. We describe a new tool, GeneMarkS-T, for ab initio identification of protein-coding regions, complete or incomplete, in RNA transcripts assembled from RNA-Seq reads. Important feature of GeneMarkS-T is unsupervised estimation of parameters of the algorithm that makes unnecessary several conventional steps used in the gene prediction protocols, most importantly the manually curated preparation of training sets. We demonstrate that i/the GeneMarkS-T self-training is robust with respect to the presence of errors in assembled transcripts and ii/accuracy of GeneMarkS-T in identification of protein-coding regions and, particularly, in prediction of gene starts compares favorably to other existing methods.

Gene annotation in viruses often relies upon similarity search methods. These methods possess high specificity but some genes may be missed, either those unique to a particular genome or those highly divergent from known homologs. To identify potentially missing viral genes we have analyzed all complete viral genomes currently available in GenBank with a specialized and augmented version of the gene finding program GeneMarkS. In particular, by implementing genome-specific self-training protocols we have better adjusted the GeneMarkS statistical models to sequences of viral genomes. Hundreds of new genes were identified, some in well studied viral genomes. For example, a new gene predicted in the genome of the Epstein-Barr virus was shown to encode a protein similar to alpha-herpesvirus minor tegument protein UL14 with heat shock functions. Convincing evidence of this similarity was obtained after only 12 PSI-BLAST iterations. In another example, several iterations of PSI-BLAST were required to demonstrate that a gene predicted in the genome of Alcelaphine herpesvirus 1 encodes a BALF1-like protein which is thought to be involved in apoptosis regulation and, potentially, carcinogenesis. New predictions were used to refine annotations of viral genomes in the RefSeq collection curated by the National Center for Biotechnology Information. Importantly, even in those cases where no sequence similarities were detected, GeneMarkS significantly reduced the number of primary targets for experimental characterization by identifying the most probable candidate genes. The new genome annotations were stored in VIOLIN, an interactive database which provides access to similarity search tools for up-to-date analysis of predicted viral proteins.

We present a new approach to automatic training of a eukaryotic ab initio gene finding algorithm. With the advent of Next-Generation Sequencing, automatic training has become paramount, allowing genome annotation pipelines to keep pace with the speed of genome sequencing. Earlier we developed GeneMark-ES, currently the only gene finding algorithm for eukaryotic genomes that performs automatic training in unsupervised ab initio mode. The new algorithm, GeneMark-ET augments GeneMark-ES with a novel method that integrates RNA-Seq read alignments into the self-training procedure. Use of ‘assembled’ RNA-Seq transcripts is far from trivial; significant error rate of assembly was revealed in recent assessments. We demonstrated in computational experiments that the proposed method of incorporation of ‘unassembled’ RNA-Seq reads improves the accuracy of gene prediction; particularly, for the 1.3 GB genome of Aedes aegypti the mean value of prediction Sensitivity and Specificity at the gene level increased over GeneMark-ES by 24.5%. In the current surge of genomic data when the need for accurate sequence annotation is higher than ever, GeneMark-ET will be a valuable addition to the narrow arsenal of automatic gene prediction tools.

Improving the accuracy of prediction of gene starts is one of a few remaining open problems in computer prediction of prokaryotic genes. Its difficulty is caused by the absence of relatively strong sequence patterns identifying true translation initiation sites. In the current paper we show that the accuracy of gene start prediction can be improved by combining models of protein-coding and non-coding regions and models of regulatory sites near gene start within an iterative Hidden Markov model based algorithm. The new gene prediction method, called GeneMarkS, utilizes a non-supervised training procedure and can be used for a newly sequenced prokaryotic genome with no prior knowledge of any protein or rRNA genes. The GeneMarkS implementation uses an improved version of the gene finding program GeneMark.hmm, heuristic Markov models of coding and non-coding regions and the Gibbs sampling multiple alignment program. GeneMarkS predicted precisely 83.2% of the translation starts of GenBank annotated Bacillus subtilis genes and 94.4% of translation starts in an experimentally validated set of Escherichia coli genes. We have also observed that GeneMarkS detects prokaryotic genes, in terms of identifying open reading frames containing real genes, with an accuracy matching the level of the best currently used gene detection methods. Accurate translation start prediction, in addition to the refinement of protein sequence N-terminal data, provides the benefit of precise positioning of the sequence region situated upstream to a gene start. Therefore, sequence motifs related to transcription and translation regulatory sites can be revealed and analyzed with higher precision. These motifs were shown to possess a significant variability, the functional and evolutionary connections of which are discussed.

Computer methods of accurate gene finding in DNA sequences require models of protein coding and non-coding regions derived either from experimentally validated training sets or from large amounts of anonymous DNA sequence. Here we propose a new, heuristic method producing fairly accurate inhomogeneous Markov models of protein coding regions. The new method needs such a small amount of DNA sequence data that the model can be built 'on the fly' by a web server for any DNA sequence >400 nt. Tests on 10 complete bacterial genomes performed with the GeneMark.hmm program demonstrated the ability of the new models to detect 93.1% of annotated genes on average, while models built by traditional training predict an average of 93.9% of genes. Models built by the heuristic approach could be used to find genes in small fragments of anonymous prokaryotic genomes and in genomes of organelles, viruses, phages and plasmids, as well as in highly inhomogeneous genomes where adjustment of models to local DNA composition is needed. The heuristic method also gives an insight into the mechanism of codon usage pattern evolution.

The number of completely sequenced bacterial genomes has been growing fast. There are computer methods available for finding genes but yet there is a need for more accurate algorithms. The GeneMark. hmm algorithm presented here was designed to improve the gene prediction quality in terms of finding exact gene boundaries. The idea was to embed the GeneMark models into naturally derived hidden Markov model framework with gene boundaries modeled as transitions between hidden states. We also used the specially derived ribosome binding site pattern to refine predictions of translation initiation codons. The algorithm was evaluated on several test sets including 10 complete bacterial genomes. It was shown that the new algorithm is significantly more accurate than GeneMark in exact gene prediction. Interestingly, the high gene finding accuracy was observed even in the case when Markov models of order zero, one and two were used. We present the analysis of false positive and false negative predictions with the caution that these categories are not precisely defined if the public database annotation is used as a control.

Accurate mapping of spliced RNA-Seq reads to genomic DNA has been known as a challenging problem. Despite significant efforts invested in developing efficient algorithms, with the human genome as a primary focus, the best solution is still not known. A recently introduced tool, TrueSight, has demonstrated better performance compared with earlier developed algorithms such as TopHat and MapSplice. To improve detection of splice junctions, TrueSight uses information on statistical patterns of nucleotide ordering in intronic and exonic DNA. This line of research led to yet another new algorithm, UnSplicer, designed for eukaryotic species with compact genomes where functional alternative splicing is likely to be dominated by splicing noise. Genome-specific parameters of the new algorithm are generated by GeneMark-ES, an ab initio gene prediction algorithm based on unsupervised training. UnSplicer shares several components with TrueSight; the difference lies in the training strategy and the classification algorithm. We tested UnSplicer on RNA-Seq data sets of Arabidopsis thaliana, Caenorhabditis elegans, Cryptococcus neoformans and Drosophila melanogaster. We have shown that splice junctions inferred by UnSplicer are in better agreement with knowledge accumulated on these well-studied genomes than predictions made by earlier developed tools.

RNA-seq has proven to be a powerful technique for transcriptome profiling based on next-generation sequencing (NGS) technologies. However, due to the short length of NGS reads, it is challenging to accurately map RNA-seq reads to splice junctions (SJs), which is a critically important step in the analysis of alternative splicing (AS) and isoform construction. In this article, we describe a new method, called TrueSight, which for the first time combines RNA-seq read mapping quality and coding potential of genomic sequences into a unified model. The model is further utilized in a machine-learning approach to precisely identify SJs. Both simulations and real data evaluations showed that TrueSight achieved higher sensitivity and specificity than other methods. We applied TrueSight to new high coverage honey bee RNA-seq data to discover novel splice forms. We found that 60.3% of honey bee multi-exon genes are alternatively spliced. By utilizing gene models improved by TrueSight, we characterized AS types in honey bee transcriptome. We believe that TrueSight will be highly useful to comprehensively study the biology of alternative splicing.