Gene Structure Prediction Research Articles

Utilizing tiling microarrays for whole‐genome analysis in plants

The recent explosion in available genome sequence data has ushered in an era in which analysis of a whole genome can be performed in a single experiment. While DNA microarrays have long been the established technology for measuring gene expression levels, standard expression arrays use relatively few probes for each gene and are typically biased toward known and predicted gene structures. Recently, with the availability of complete genome sequences for many organisms, very-high-density oligonucleotide-based microarrays that span the entire genome have emerged as the preferred platform for genomic analysis. Whole-genome tiling microarrays can be employed for a myriad of purposes, including empirical annotation of the transcriptome, chromatin immunoprecipitation-chip studies, analysis of alternative RNA splicing, characterization of the methylation state of cytosine bases throughout a genome (methylome), and DNA polymorphism discovery. Here, we review several applications of whole-genome technology to obtain a variety of genomic-scale information in plants.

HMMEditor: a visual editing tool for profile hidden Markov model

BackgroundProfile Hidden Markov Model (HMM) is a powerful statistical model to represent a family of DNA, RNA, and protein sequences. Profile HMM has been widely used in bioinformatics research such as sequence alignment, gene structure prediction, motif identification, protein structure prediction, and biological database search. However, few comprehensive, visual editing tools for profile HMM are publicly available.ResultsWe develop a visual editor for profile Hidden Markov Models (HMMEditor). HMMEditor can visualize the profile HMM architecture, transition probabilities, and emission probabilities. Moreover, it provides functions to edit and save HMM and parameters. Furthermore, HMMEditor allows users to align a sequence against the profile HMM and to visualize the corresponding Viterbi path.ConclusionHMMEditor provides a set of unique functions to visualize and edit a profile HMM. It is a useful tool for biological sequence analysis and modeling. Both HMMEditor software and web service are freely available.

Using N-SCAN or TWINSCAN to predict gene structures in genomic DNA sequences.

N-SCAN is a gene-prediction system that combines the methods of ab initio predictors like GENSCAN with information derived from genome comparison. It is the latest in the TWINSCAN series of programs. This unit describes the use of N-SCAN to identify gene structures in eukaryotic genomic sequences. Protocols for using N-SCAN through its Web interface and from the command line in a Linux environment are provided. Detailed discussion about the appropriate parameter settings, input-sequence processing, and choice of genome for comparison are included.

A manually curated database of tetrapod mitochondrially encoded tRNA sequences and secondary structures

BackgroundMitochondrial tRNAs have been the subject of study for structural biologists interested in their secondary structure characteristics, evolutionary biologists have researched patterns of compensatory and structural evolution and medical studies have been directed towards understanding the basis of human disease. However, an up to date, manually curated database of mitochondrially encoded tRNAs from higher animals is currently not available.DescriptionWe obtained the complete mitochondrial sequence for 277 tetrapod species from GenBank and re-annotated all of the tRNAs based on a multiple alignment of each tRNA gene and secondary structure prediction made independently for each tRNA. The mitochondrial (mt) tRNA sequences and the secondary structure based multiple alignments are freely available as Supplemental Information online.ConclusionWe compiled a manually curated database of mitochondrially encoded tRNAs from tetrapods with completely sequenced genomes. In the course of our work, we reannotated more than 10% of all tetrapod mt-tRNAs and subsequently predicted the secondary structures of 6060 mitochondrial tRNAs. This carefully constructed database can be utilized to enhance our knowledge in several different fields including the evolution of mt-tRNA secondary structure and prediction of pathogenic mt-tRNA mutations. In addition, researchers reporting novel mitochondrial genome sequences should check their tRNA gene annotations against our database to ensure a higher level of fidelity of their annotation.

How does eukaryotic gene prediction work?

Computational prediction of gene structure is crucial for interpreting genomic sequences. But how do the algorithms involved work and how accurate are they?

2008 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology

This symposium will bring together top researchers, practitioners, and students from around the world to discuss the latest advances in the field of Computational Intelligence and its application to real world problems in biology. Computational Intelligence includes artificial neural nets, fuzzy logic, evolutionary algorithms, and other emerging techniques. Topics of interest include but are not limited to: gene expression array analysis, structure prediction and folding, drug design, molecular sequence alignment and analysis, metabolic pathway analysis, medical imaging and pattern recognition, analysis of large biological data sets, and a variety of other topics. The use of computational intelligence must play a substantial role in submitted papers. Submissions will be peer reviewed and accepted papers will be published in the conference proceedings. Paper Submission: Prospective authors are invited to submit papers of no more than eight (8) pages including results, figures and references; submission details can be found on the symposium web site: www.cibcb.org. Authors are expected to present their paper at the symposium. Special Sessions and Tutorials: It is anticipated that the symposium will have a number of special sessions and tutorials. Consult the symposium web site for instructions on proposing a special session or tutorial. Papers for special sessions must have significant computational intelligence content as do regular session papers. Location: Sun Valley is in the mountains of Central Idaho. The area is well known for its outdoor activity offerings. Whitewater rafting, mountain biking, horseback riding, fly fishing, and golf are available. Numerous shopping and entertainment possibilities also exist.

Genome-wide identification of spliced introns using a tiling microarray

The prediction of gene models from genome sequence remains an unsolved problem. One hallmark of eukaryotic gene structure is the presence of introns, which are spliced out of pre-mRNAs prior to translation. The excised introns are released in the form of lariats, which must be debranched prior to their turnover. In the yeast Saccharomyces cerevisiae, the absence of the debranching enzyme causes these lariat RNAs to accumulate. This accumulation allows a comparison of tiling array signals of RNA from the debranching mutant to the wild-type parent strain, and thus the identification of lariats on a genome-wide scale. This approach identified 141 of 272 known introns, confirmed three previously predicted introns, predicted four novel introns (of which two were experimentally confirmed), and led to the reannotation of four others. In many instances, signals from the tiling array delineated the 5' splice site and branchpoint site, confirming predicted gene structures. Nearly all introns that went undetected are present in mRNAs expressed at low levels. Overall, 97% of the significant probes could be attributed either to spliced introns or to genes up-regulated by deletion of the debranching enzyme. Because the debranching enzyme is conserved among eukaryotes, this approach could be generally applicable for the annotation of eukaryotic genes and the detection of novel and alternative splice forms.

Analysis of rice glycosyl hydrolase family 1 and expression of Os4bglu12 β-glucosidase

BackgroundGlycosyl hydrolase family 1 (GH1) β-glucosidases have been implicated in physiologically important processes in plants, such as response to biotic and abiotic stresses, defense against herbivores, activation of phytohormones, lignification, and cell wall remodeling. Plant GH1 β-glucosidases are encoded by a multigene family, so we predicted the structures of the genes and the properties of their protein products, and characterized their phylogenetic relationship to other plant GH1 members, their expression and the activity of one of them, to begin to decipher their roles in rice.ResultsForty GH1 genes could be identified in rice databases, including 2 possible endophyte genes, 2 likely pseudogenes, 2 gene fragments, and 34 apparently competent rice glycosidase genes. Phylogenetic analysis revealed that GH1 members with closely related sequences have similar gene structures and are often clustered together on the same chromosome. Most of the genes appear to have been derived from duplications that occurred after the divergence of rice and Arabidopsis thaliana lineages from their common ancestor, and the two plants share only 8 common gene lineages. At least 31 GH1 genes are expressed in a range of organs and stages of rice, based on the cDNA and EST sequences in public databases. The cDNA of the Os4bglu12 gene, which encodes a protein identical at 40 of 44 amino acid residues with the N-terminal sequence of a cell wall-bound enzyme previously purified from germinating rice, was isolated by RT-PCR from rice seedlings. A thioredoxin-Os4bglu12 fusion protein expressed in Escherichia coli efficiently hydrolyzed β-(1,4)-linked oligosaccharides of 3–6 glucose residues and laminaribiose.ConclusionCareful analysis of the database sequences produced more reliable rice GH1 gene structure and protein product predictions. Since most of these genes diverged after the divergence of the ancestors of rice and Arabidopsis thaliana, only a few of their functions could be implied from those of GH1 enzymes from Arabidopsis and other dicots. This implies that analysis of GH1 enzymes in monocots is necessary to understand their function in the major grain crops. To begin this analysis, Os4bglu12 β-glucosidase was characterized and found to have high exoglucanase activity, consistent with a role in cell wall metabolism.

Evolutionary computation in bioinformatics: a review

This paper provides an overview of the application of evolutionary algorithms in certain bioinformatics tasks. Different tasks such as gene sequence analysis, gene mapping, deoxyribonucleic acid (DNA) fragment assembly, gene finding, microarray analysis, gene regulatory network analysis, phylogenetic trees, structure prediction and analysis of DNA, ribonucleic acid and protein, and molecular docking with ligand design are, first of all, described along with their basic features. The relevance of using evolutionary algorithms to these problems is then mentioned. These are followed by different approaches, along with their merits, for addressing some of the aforesaid tasks. Finally, some limitations of the current research activity are provided. An extensive bibliography is included

Prediction of Genomic Functional Elements

As the number of sequenced genomes increases, the ability to deduce genome function becomes increasingly salient. For many genome sequences, the only annotation that will be available for the foreseeable future will be based on computational predictions and comparisons with functional elements in related species. Here we discuss computational approaches for automated genome-wide annotation of functional elements in mammalian genomes. These include methods for ab initio and comparative gene-structure predictions. Gene features such as intron splice sites, 3' untranslated regions, promoters, and cis-regulatory elements are discussed, as is a novel method for predicting DNaseI hypersensitive sites. Recent methodologies for predicting noncoding RNA genes, including microRNA genes and their targets, are also reviewed.

Contemporary Progress in Gene Structure Prediction

Contemporary Progress in Gene Structure Prediction

Using ESTs to improve the accuracy of de novo gene prediction.

BackgroundESTs are a tremendous resource for determining the exon-intron structures of genes, but even extensive EST sequencing tends to leave many exons and genes untouched. Gene prediction systems based exclusively on EST alignments miss these exons and genes, leading to poor sensitivity. De novo gene prediction systems, which ignore ESTs in favor of genomic sequence, can predict such "untouched" exons, but they are less accurate when predicting exons to which ESTs align. TWINSCAN is the most accurate de novo gene finder available for nematodes and N-SCAN is the most accurate for mammals, as measured by exact CDS gene prediction and exact exon prediction.ResultsTWINSCAN_EST is a new system that successfully combines EST alignments with TWINSCAN. On the whole C. elegans genome TWINSCAN_EST shows 14% improvement in sensitivity and 13% in specificity in predicting exact gene structures compared to TWINSCAN without EST alignments. Not only are the structures revealed by EST alignments predicted correctly, but these also constrain the predictions without alignments, improving their accuracy. For the human genome, we used the same approach with N-SCAN, creating N-SCAN_EST. On the whole genome, N-SCAN_EST produced a 6% improvement in sensitivity and 1% in specificity of exact gene structure predictions compared to N-SCAN.ConclusionTWINSCAN_EST and N-SCAN_EST are more accurate than TWINSCAN and N-SCAN, while retaining their ability to discover novel genes to which no ESTs align. Thus, we recommend using the EST versions of these programs to annotate any genome for which EST information is available.TWINSCAN_EST and N-SCAN_EST are part of the TWINSCAN open source software package .

On the Complexity of Sparse Exon Assembly

Gene structure prediction is one of the most important problems in computational molecular biology. It involves two steps: the first is finding the evidence (e.g., predicting splice sites) and the second is interpreting the evidence, that is, trying to determine the whole gene structure by assembling its pieces. In this paper, we suggest a combinatorial solution to the second step, which is also referred to as the "Exon Assembly Problem." We use a similarity-based approach that aims to produce a single gene structure based on similarities to a known homologous sequence. We target the sparse case, where filtering has been applied to the data, resulting in a set of O(n) candidate exon blocks. Our algorithm yields an O(n(2) square root of n) solution.

Bioinformatics Resources for Cancer Research with an Emphasis on Gene Function and Structure Prediction Tools

The immensely popular fields of cancer research and bioinformatics overlap in many different areas, e.g. large data repositories that allow for users to analyze data from many experiments (data handling, databases), pattern mining, microarray data analysis, and interpretation of proteomics data. There are many newly available resources in these areas that may be unfamiliar to most cancer researchers wanting to incorporate bioinformatics tools and analyses into their work, and also to bioinformaticians looking for real data to develop and test algorithms. This review reveals the interdependence of cancer research and bioinformatics, and highlight the most appropriate and useful resources available to cancer researchers. These include not only public databases, but general and specific bioinformatics tools which can be useful to the cancer researcher. The primary foci are function and structure prediction tools of protein genes. The result is a useful reference to cancer researchers and bioinformaticians studying cancer alike.

Vertebrate gene finding from multiple-species alignments using a two-level strategy

BackgroundOne way in which the accuracy of gene structure prediction in vertebrate DNA sequences can be improved is by analyzing alignments with multiple related species, since functional regions of genes tend to be more conserved.ResultsWe describe DOGFISH, a vertebrate gene finder consisting of a cleanly separated site classifier and structure predictor. The classifier scores potential splice sites and other features, using sequence alignments between multiple vertebrate species, while the structure predictor hypothesizes coding transcripts by combining these scores using a simple model of gene structure. This also identifies and assigns confidence scores to possible additional exons. Performance is assessed on the ENCODE regions. We predict transcripts and exons across the whole human genome, and identify over 10,000 high confidence new coding exons not in the Ensembl gene set.ConclusionWe present a practical multiple species gene prediction method. Accuracy improves as additional species, up to at least eight, are introduced. The novel predictions of the whole-genome scan should support efficient experimental verification.

Managing Evidence from Multiple Gene Finding Resources via an XML-based Integration Architecture

SummaryWhile biological processes underlying gene expression are still under experimental research, computational gene prediction techniques have reached high level of sophistication with the employment of efficient intrinsic and extrinsic methods that identify protein-coding regions within query genomic sequences. Their ability though to delineate the exact exon boundaries is characterized by a trade off between sensitivity and specificity and still is prone to alternations in gene regulation during transcription and splicing and to inherent complexities introduced by the implemented methodology. Evaluation studies have shown that combinatorial approaches exhibit improved accuracy levels through the integration of evidence data from multiple resources that are further assessed in order to end up with the most probable gene assembly.In this work, we present an integration and information handling architecture that exploits evidence derived from multiple gene finding resources, in order to generate machine-readable representations of optimal/suboptimal gene structure predictions, signal features identification and high scoring similarity matches. Unlike most combinatorial techniques, which end up with the most probable gene assembly, the objective of this architecture is to support advanced information handling mechanisms that may give more in depth insights on the underlying gene expression machinery and the alternations that may occur. Technically, XML was adopted to build and interchange structured data among the architecture’s components together with relevant technologies offering graphical representations and queries formulation/execution over single/multiple information sources.

Reevaluation of the Predicted Gene Structure of Dictyostelium Cystatin A3 (cpiC) by Nucleotide Sequence Determination of its cDNA* and its Phylogenetic Position in the Cystatin Superfamily

Cystatins, cysteine protease inhibitors, are widely distributed among eukaryotes. We reevaluated the structure of the gene cpiC, a gene encoding the third identified member of cystatin family (cystatin A3) that was predicted in the genome database of the social amoeba Dictyostelium discoidium (dictyBase) but remained controversial. We determined the sequences of cDNA and PCR-amplified genomic DNA fragment and found a critical error in the registered nucleotide sequence. The corrected cystatin A3 gene has an open reading frame (ORF) without intron sequence interruption and encodes 94 amino acids (aa), in contrast to the previously predicted sequence of either 80, 82 or 118 aa. The cDNA has an unusual internal poly(A) sequence of 31 adenines, which immediately follows the translation termination codon (TAA) located 146 nucleotides upstream of the post-transcriptional polyadenylation site. The amino acid sequence of Dictyostelium cystatin A3 shows a high similarity to those of previously reported Dictyostelium cystatins as well as Family I cystatins of higher eukaryotes.

Engineering a software tool for gene structure prediction in higher organisms

The research area now commonly called ‘bioinformatics’ has brought together biologists, computer scientists, statisticians, and scientists of many other fields of expertise to work on computational solutions to biological problems. A large number of algorithms and software packages are freely available for many specific tasks, such as sequence alignment, molecular phylogeny reconstruction, or protein structure determination. Rapidly changing needs and demands on data handling capacity challenge the application providers to consistently keep pace. In practice, this has led to many incremental advances and re-writing of code that present the user community with confusing options and a large overhead from non-standardized implementations that need to be integrated into existing work flows. This situation gives much scope for contributions by software engineers. In this article, we describe an example of engineering a software tool for a specific bioinformatics task known as spliced alignment. The problem was motivated by disabling limitations in an original, ad hoc, and yet widely popular implementation by one of the authors. The present collaboration has led to a robust, highly versatile, and extensible tool (named GenomeThreader) that not only overcomes the limitations of the earlier implementation but greatly improves space and time requirements.

Incorporation of splice site probability models for non-canonical introns improves gene structure prediction in plants

The vast majority of introns in protein-coding genes of higher eukaryotes have a GT dinucleotide at their 5'-terminus and an AG dinucleotide at their 3' end. About 1-2% of introns are non-canonical, with the most abundant subtype of non-canonical introns being characterized by GC and AG dinucleotides at their 5'- and 3'-termini, respectively. Most current gene prediction software, whether based on ab initio or spliced alignment approaches, does not include explicit models for non-canonical introns or may exclude their prediction altogether. With present amounts of genome and transcript data, it is now possible to apply statistical methodology to non-canonical splice site prediction. We pursued one such approach and describe the training and implementation of GC-donor splice site models for Arabidopsis and rice, with the goal of exploring whether specific modeling of non-canonical introns can enhance gene structure prediction accuracy. Our results indicate that the incorporation of non-canonical splice site models yields dramatic improvements in annotating genes containing GC-AG and AT-AC non-canonical introns. Comparison of models shows differences between monocot and dicot species, but also suggests GC intron-specific biases independent of taxonomic clade. We also present evidence that GC-AG introns occur preferentially in genes with atypically high exon counts. Source code for the updated versions of GeneSeqer and SplicePredictor (distributed with the GeneSeqer code) isavailable at http://bioinformatics.iastate.edu/bioinformatics2go/gs/download.html. Web servers for Arabidopsis, rice and other plant species are accessible at http://www.plantgdb.org/PlantGDB-cgi/GeneSeqer/AtGDBgs.cgi, http://www.plantgdb.org/PlantGDB-cgi/GeneSeqer/OsGDBgs.cgi and http://www.plantgdb.org/PlantGDB-cgi/GeneSeqer/PlantGDBgs.cgi, respectively. A SplicePredictor web server is available at http://bioinformatics.iastate.edu/cgi-bin/sp.cgi. Software to generate training data and parameterizations for Bayesian splice site models is available at http://gremlin1.gdcb.iastate.edu/~volker/SB05B/BSSM4GSQ/

A method of precise mRNA/DNA homology-based gene structure prediction

BackgroundAccurate and automatic gene finding and structural prediction is a common problem in bioinformatics, and applications need to be capable of handling non-canonical splice sites, micro-exons and partial gene structure predictions that span across several genomic clones.ResultsWe present a mRNA/DNA homology based gene structure prediction tool, GIGOgene. We use a new affine gap penalty splice-enhanced global alignment algorithm running in linear memory for a high quality annotation of splice sites. Our tool includes a novel algorithm to assemble partial gene structure predictions using interval graphs. GIGOgene exhibited a sensitivity of 99.08% and a specificity of 99.98% on the Genie learning set, and demonstrated a higher quality of gene structural prediction when compared to Sim4, est2genome, Spidey, Galahad and BLAT, including when genes contained micro-exons and non-canonical splice sites. GIGOgene showed an acceptable loss of prediction quality when confronted with a noisy Genie learning set simulating ESTs.ConclusionGIGOgene shows a higher quality of gene structure prediction for mRNA/DNA spliced alignment when compared to other available tools.