Performance assessment of promoter predictions on ENCODE regions in the EGASP experiment

Deep learning and support vector machines for transcription start site identification

Recognizing transcription start sites is key to gene identification. Several approaches have been employed in related problems such as detecting translation initiation sites or promoters, many of the most recent ones based on machine learning. Deep learning methods have been proven to be exceptionally effective for this task, but their use in transcription start site identification has not yet been explored in depth. Also, the very few existing works do not compare their methods to support vector machines (SVMs), the most established technique in this area of study, nor provide the curated dataset used in the study. The reduced amount of published papers in this specific problem could be explained by this lack of datasets. Given that both support vector machines and deep neural networks have been applied in related problems with remarkable results, we compared their performance in transcription start site predictions, concluding that SVMs are computationally much slower, and deep learning methods, specially long short-term memory neural networks (LSTMs), are best suited to work with sequences than SVMs. For such a purpose, we used the reference human genome GRCh38. Additionally, we studied two different aspects related to data processing: the proper way to generate training examples and the imbalanced nature of the data. Furthermore, the generalization performance of the models studied was also tested using the mouse genome, where the LSTM neural network stood out from the rest of the algorithms. To sum up, this article provides an analysis of the best architecture choices in transcription start site identification, as well as a method to generate transcription start site datasets including negative instances on any species available in Ensembl. We found that deep learning methods are better suited than SVMs to solve this problem, being more efficient and better adapted to long sequences and large amounts of data. We also create a transcription start site (TSS) dataset large enough to be used in deep learning experiments.

BERT-Promoter: An improved sequence-based predictor of DNA promoter using BERT pre-trained model and SHAP feature selection

A promoter is a sequence of DNA that initializes the process of transcription and regulates whenever and wherever genes are expressed in the organism. Because of its importance in molecular biology, identifying DNA promoters are challenging to provide useful information related to its functions and related diseases. Several computational models have been developed to early predict promoters from high-throughput sequencing over the past decade. Although some useful predictors have been proposed, there remains short-falls in those models and there is an urgent need to enhance the predictive performance to meet the practice requirements. In this study, we proposed a novel architecture that incorporated transformer natural language processing (NLP) and explainable machine learning to address this problem. More specifically, a pre-trained Bidirectional Encoder Representations from Transformers (BERT) model was employed to encode DNA sequences, and SHapley Additive exPlanations (SHAP) analysis served as a feature selection step to look at the top-rank BERT encodings. At the last stage, different machine learning classifiers were implemented to learn the top features and produce the prediction outcomes. This study not only predicted the DNA promoters but also their activities (strong or weak promoters). Overall, several experiments showed an accuracy of 85.5 % and 76.9 % for these two levels, respectively. Our performance showed a superiority to previously published predictors on the same dataset in most measurement metrics. We named our predictor as BERT-Promoter and it is freely available at https://github.com/khanhlee/bert-promoter.

Supervised promoter recognition: a benchmark framework

Motivation

Deep learning has become a prevalent method in identifying genomic regulatory sequences such as promoters. In a number of recent papers, the performance of deep learning models has continually been reported as an improvement over alternatives for sequence-based promoter recognition. However, the performance improvements in these models do not account for the different datasets that models are evaluated on. The lack of a consensus dataset and procedure for benchmarking purposes has made the comparison of each model’s true performance difficult to assess.

Results

We present a framework called Supervised Promoter Recognition Framework (‘SUPR REF’) capable of streamlining the complete process of training, validating, testing, and comparing promoter recognition models in a systematic manner. SUPR REF includes the creation of biologically relevant benchmark datasets to be used in the evaluation process of deep learning promoter recognition models. We showcase this framework by comparing the models’ performances on alternative datasets, and properly evaluate previously published models on new benchmark datasets. Our results show that the reliability of deep learning ab initio promoter recognition models on eukaryotic genomic sequences is still not at a sufficient level, as overall performance is still low. These results originate from a subset of promoters, the well-known RNA Polymerase II core promoters. Furthermore, given the observational nature of these data, cross-validation results from small promoter datasets need to be interpreted with caution.

Identification of potential MTAs and candidate genes for juice quality- and yield-related traits in Saccharum clones: a genome-wide association and comparative genomic study

Sugarcane is an economically important commercial crop which provides raw material for the production of sugar, jaggery, bioethanol, biomass and other by-products. Sugarcane breeding till today heavily relies on conventional breeding approaches which is time consuming, laborious and costly. Integration of marker-assisted selection (MAS) in sugarcane genetic improvement programs for difficult to select traits like sucrose content, resistance to pests and diseases and tolerance to abiotic stresses will accelerate varietal development. In the present study, association mapping approach was used to identify QTLs and genes associated with sucrose and other important yield-contributing traits. A mapping panel of 110 diverse sugarcane genotypes and 148 microsatellite primers were used for structured association mapping study. An optimal subpopulation number (ΔK) of 5 was identified by structure analysis. GWAS analysis using TASSEL identified a total of 110 MTAs which were localized into 27 QTLs by GLM and MLM (Q + K, PC + K) approaches. Among the 24 QTLs sequenced, 12 were able to identify potential candidate genes, viz., starch branching enzyme, starch synthase 4, sugar transporters and G3P-DH related to carbohydrate metabolism and hormone pathway-related genes ethylene insensitive 3-like 1, reversion to ethylene sensitive1-like, and auxin response factor associated to juice quality- and yield-related traits. Six markers, NKS 5_185, SCB 270_144, SCB 370_256, NKS 46_176 and UGSM 648_245, associated with juice quality traits and marker SMC31CUQ_304 associated with NMC were validated and identified as significantly associated to the traits by one-way ANOVA analysis. In conclusion, 24 potential QTLs identified in the present study could be used in sugarcane breeding programs after further validation in larger population. The candidate genes from carbohydrate and hormone response pathway presented in this study could be manipulated with genome editing approaches to further improve sugarcane crop.

Critical assessment of computational tools for prokaryotic and eukaryotic promoter prediction

Promoters are crucial regulatory DNA regions for gene transcriptional activation. Rapid advances in next-generation sequencing technologies have accelerated the accumulation of genome sequences, providing increased training data to inform computational approaches for both prokaryotic and eukaryotic promoter prediction. However, it remains a significant challenge to accurately identify species-specific promoter sequences using computational approaches. To advance computational support for promoter prediction, in this study, we curated 58 comprehensive, up-to-date, benchmark datasets for 7 different species (i.e. Escherichia coli, Bacillus subtilis, Homo sapiens, Mus musculus, Arabidopsis thaliana, Zea mays and Drosophila melanogaster) to assist the research community to assess the relative functionality of alternative approaches and support future research on both prokaryotic and eukaryotic promoters. We revisited 106 predictors published since 2000 for promoter identification (40 for prokaryotic promoter, 61 for eukaryotic promoter, and 5 for both). We systematically evaluated their training datasets, computational methodologies, calculated features, performance and software usability. On the basis of these benchmark datasets, we benchmarked 19 predictors with functioning webservers/local tools and assessed their prediction performance. We found that deep learning and traditional machine learning-based approaches generally outperformed scoring function-based approaches. Taken together, the curated benchmark dataset repository and the benchmarking analysis in this study serve to inform the design and implementation of computational approaches for promoter prediction and facilitate more rigorous comparison of new techniques in the future.

Allelic variation in sugary1 gene affecting kernel sweetness among diverse-mutant and -wild-type maize inbreds

Sweet corn is popular worldwide as vegetable. Though large numbers of sugary1 (su1)-based sweet corn germplasm are available, allelic diversity in su1 gene encoding SU1 isoamylase among diverse maize inbreds has not been analyzed. Here, we characterized the su1 gene in maize and compared with allied species. The entire su1 gene (11,720 bp) was sequenced among six mutant (su1) and five wild (Su1) maize inbreds. Fifteen InDels of 2-45 bp were selected to develop markers for studying allelic diversity in su1 gene among 19 mutant- (su1) and 29 wild-type (Su1) inbreds. PIC ranged from 0.15 (SU-InDel7) to 0.37 (SU-InDel13). Major allele frequency varied from 0.52 to 0.90, while gene diversity ranged from 0.16 to 0.49. Phylogenetic tree categorized 48 maize inbreds in two clusters each for wild- type (Su1) and mutant (su1) types. 44 haplotypes of su1 were observed, with three haplotypes (Hap6, Hap22 and Hap29) sharing more than one genotype. Further, comparisons were made with 23 orthologues of su1 from 16 grasses and Arabidopsis. Maize possessed 15-19 exons in su1, while it was 11-24 exons among orthologues. Introns among the orthologues were longer (77-2206 bp) than maize (859-1718 bp). SU1 protein of maize and orthologues had conserved α-amylase and CBM_48 domains. The study also provided physicochemical properties and secondary structure of SU1 protein in maize and its orthologues. Phylogenetic analysis showed closer relationship of maize SU1 protein with P. hallii, S. bicolor and E. tef than Triticum sp. and Oryza sp. The study showed that presence of high allelic diversity in su1 gene which can be utilized in the sweet corn breeding program. This is the first report of comprehensive characterization of su1 gene and its allelic forms in diverse maize and related orthologues.

A prescient evolutionary model for genesis, duplication and differentiation of MIR160 homologs in Brassicaceae

MicroRNA160 is a class of nitrogen-starvation responsive genes which governs establishment of root system architecture by down-regulating AUXIN RESPONSE FACTOR genes (ARF10, ARF16 and ARF17) in plants. The high copy number of MIR160 variants discovered by us from land plants, especially polyploid crop Brassicas, posed questions regarding genesis, duplication, evolution and function. Absence of studies on impact of whole genome and segmental duplication on retention and evolution of MIR160 homologs in descendent plant lineages prompted us to undertake the current study. Herein, we describe ancestry and fate of MIR160 homologs in Brassicaceae in context of polyploidy driven genome re-organization, copy number and differentiation. Paralogy amongst Brassicaceae MIR160a, MIR160b and MIR160c was inferred using phylogenetic analysis of 468 MIR160 homologs from land plants. The evolutionarily distinct MIR160a was found to represent ancestral form and progenitor of MIR160b and MIR160c. Chronology of evolutionary events resulting in origin and diversification of genomic loci containing MIR160 homologs was delineated using derivatives of comparative synteny. A prescient model for causality of segmental duplications in establishment of paralogy in Brassicaceae MIR160, with whole genome duplication accentuating the copy number increase, is being posited in which post-segmental duplication events viz. differential gene fractionation, gene duplications and inversions are shown to drive divergence of chromosome segments. While mutations caused the diversification of MIR160a, MIR160b and MIR160c, duplicated segments containing these diversified genes suffered gene rearrangements via gene loss, duplications and inversions. Yet the topology of phylogenetic and phenetic trees were found congruent suggesting similar evolutionary trajectory. Over 80% of Brassicaceae genomes and subgenomes showed a preferential retention of single copy each of MIR160a, MIR160b and MIR160c suggesting functional relevance. Thus, our study provides a blue-print for reconstructing ancestry and phylogeny of MIRNA gene families at genomics level and analyzing the impact of polyploidy on organismal complexity. Such studies are critical for understanding the molecular basis of agronomic traits and deploying appropriate candidates for crop improvement.

Computational Prediction of Sigma-54 Promoters in Bacterial Genomes by Integrating Motif Finding and Machine Learning Strategies

Sigma factor, as a unit of RNA polymerase holoenzyme, is a critical factor in the process of gene transcriptional regulation. It recognizes the specific DNA sites and brings the core enzyme of RNA polymerase to the upstream regions of target genes. Therefore, the prediction of the promoters for a particular sigma factor is essential for interpreting functional genomic data and observation. This paper develops a new method to predict sigma-54 promoters in bacterial genomes. The new method organically integrates motif finding and machine learning strategies to capture the intrinsic features of sigma-54 promoters. The experiments on E. coli benchmark test set show that our method has good capability to distinguish sigma-54 promoters from surrounding or randomly selected DNA sequences. The applications of the other three bacterial genomes indicate the potential robustness and applicative power of our method on a large number of bacterial genomes. The source code of our method can be freely downloaded at https://github.com/maqin2001/PromotePredictor.

Promoter analysis and prediction in the human genome using sequence-based deep learning models

Motivation

Computational identification of promoters is notoriously difficult as human genes often have unique promoter sequences that provide regulation of transcription and interaction with transcription initiation complex. While there are many attempts to develop computational promoter identification methods, we have no reliable tool to analyze long genomic sequences.

Results

In this work, we further develop our deep learning approach that was relatively successful to discriminate short promoter and non-promoter sequences. Instead of focusing on the classification accuracy, in this work we predict the exact positions of the transcription start site inside the genomic sequences testing every possible location. We studied human promoters to find effective regions for discrimination and built corresponding deep learning models. These models use adaptively constructed negative set, which iteratively improves the model’s discriminative ability. Our method significantly outperforms the previously developed promoter prediction programs by considerably reducing the number of false-positive predictions. We have achieved error-per-1000-bp rate of 0.02 and have 0.31 errors per correct prediction, which is significantly better than the results of other human promoter predictors.

Availability and implementation

The developed method is available as a web server at http://www.cbrc.kaust.edu.sa/PromID/.

Nucleotide patterns aiding in prediction of eukaryotic promoters

Computational analysis of promoters is hindered by the complexity of their architecture. In less studied genomes with complex organization, false positive promoter predictions are common. Accurate identification of transcription start sites and core promoter regions remains an unsolved problem. In this paper, we present a comprehensive analysis of genomic features associated with promoters and show that probabilistic integrative algorithms-driven models allow accurate classification of DNA sequence into “promoters” and “non-promoters” even in absence of the full-length cDNA sequences. These models may be built upon the maps of the distributions of sequence polymorphisms, RNA sequencing reads on genomic DNA, methylated nucleotides, transcription factor binding sites, as well as relative frequencies of nucleotides and their combinations. Positional clustering of binding sites shows that the cells of Oryza sativa utilize three distinct classes of transcription factors: those that bind preferentially to the [-500,0] region (188 “promoter-specific” transcription factors), those that bind preferentially to the [0,500] region (282 “5′ UTR-specific” TFs), and 207 of the “promiscuous” transcription factors with little or no location preference with respect to TSS. For the most informative motifs, their positional preferences are conserved between dicots and monocots.

Recognition of prokaryotic and eukaryotic promoters using convolutional deep learning neural networks

Accurate computational identification of promoters remains a challenge as these key DNA regulatory regions have variable structures composed of functional motifs that provide gene-specific initiation of transcription. In this paper we utilize Convolutional Neural Networks (CNN) to analyze sequence characteristics of prokaryotic and eukaryotic promoters and build their predictive models. We trained a similar CNN architecture on promoters of five distant organisms: human, mouse, plant (Arabidopsis), and two bacteria (Escherichia coli and Bacillus subtilis). We found that CNN trained on sigma70 subclass of Escherichia coli promoter gives an excellent classification of promoters and non-promoter sequences (Sn = 0.90, Sp = 0.96, CC = 0.84). The Bacillus subtilis promoters identification CNN model achieves Sn = 0.91, Sp = 0.95, and CC = 0.86. For human, mouse and Arabidopsis promoters we employed CNNs for identification of two well-known promoter classes (TATA and non-TATA promoters). CNN models nicely recognize these complex functional regions. For human promoters Sn/Sp/CC accuracy of prediction reached 0.95/0.98/0,90 on TATA and 0.90/0.98/0.89 for non-TATA promoter sequences, respectively. For Arabidopsis we observed Sn/Sp/CC 0.95/0.97/0.91 (TATA) and 0.94/0.94/0.86 (non-TATA) promoters. Thus, the developed CNN models, implemented in CNNProm program, demonstrated the ability of deep learning approach to grasp complex promoter sequence characteristics and achieve significantly higher accuracy compared to the previously developed promoter prediction programs. We also propose random substitution procedure to discover positionally conserved promoter functional elements. As the suggested approach does not require knowledge of any specific promoter features, it can be easily extended to identify promoters and other complex functional regions in sequences of many other and especially newly sequenced genomes. The CNNProm program is available to run at web server http://www.softberry.com.

SeqTools: visual tools for manual analysis of sequence alignments

Background

Manual annotation is essential to create high-quality reference alignments and annotation. Annotators need to be able to view sequence alignments in detail. The SeqTools package provides three tools for viewing different types of sequence alignment: Blixem is a many-to-one browser of pairwise alignments, displaying multiple match sequences aligned against a single reference sequence; Dotter provides a graphical dot-plot view of a single pairwise alignment; and Belvu is a multiple sequence alignment viewer, editor, and phylogenetic tool. These tools were originally part of the AceDB genome database system but have been completely rewritten to make them generally available as a standalone package of greatly improved function.

Findings

Blixem is used by annotators to give a detailed view of the evidence for particular gene models. Blixem displays the gene model positions and the match sequences aligned against the genomic reference sequence. Annotators use this for many reasons, including to check the quality of an alignment, to find missing/misaligned sequence and to identify splice sites and polyA sites and signals. Dotter is used to give a dot-plot representation of a particular pairwise alignment. This is used to identify sequence that is not represented (or is misrepresented) and to quickly compare annotated gene models with transcriptional and protein evidence that putatively supports them. Belvu is used to analyse conservation patterns in multiple sequence alignments and to perform a combination of manual and automatic processing of the alignment. High-quality reference alignments are essential if they are to be used as a starting point for further automatic alignment generation.

Conclusions

While there are many different alignment tools available, the SeqTools package provides unique functionality that annotators have found to be essential for analysing sequence alignments as part of the manual annotation process.

Electronic supplementary material

The online version of this article (doi:10.1186/s13104-016-1847-3) contains supplementary material, which is available to authorized users.

Mesoscopic model and free energy landscape for protein-DNA binding sites: analysis of cyanobacterial promoters

The identification of protein binding sites in promoter sequences is a key problem to understand and control regulation in biochemistry and biotechnological processes. We use a computational method to analyze promoters from a given genome. Our approach is based on a physical model at the mesoscopic level of protein-DNA interaction based on the influence of DNA local conformation on the dynamics of a general particle along the chain. Following the proposed model, the joined dynamics of the protein particle and the DNA portion of interest, only characterized by its base pair sequence, is simulated. The simulation output is analyzed by generating and analyzing the Free Energy Landscape of the system. In order to prove the capacity of prediction of our computational method we have analyzed nine promoters of Anabaena PCC 7120. We are able to identify the transcription starting site of each of the promoters as the most populated macrostate in the dynamics. The developed procedure allows also to characterize promoter macrostates in terms of thermo-statistical magnitudes (free energy and entropy), with valuable biological implications. Our results agree with independent previous experimental results. Thus, our methods appear as a powerful complementary tool for identifying protein binding sites in promoter sequences.

Binding of specific proteins to particular sites in the DNA sequence is a fundamental issue for gene regulation in molecular biology and genetic engineering. A deep understanding of cell physiology requires the analysis of a plethora of genes involving characterization of their promoter architectures that determine their regulation and gene transcription. In order to locate the promoter elements of a given gene, experimental determination of its transcription start site (TSS) is required. This is an expensive, time-consuming task that, depending on our requirements, could be simplified using computational analysis as a first approach. Nevertheless, most computational methods lack a physical basis on the protein-DNA interaction mechanism. We adopt here this strategy, by using a simple model for protein-DNA interaction to find TSS in a bunch of cyanobacteria promoters. We make use of physical tools to characterize these TSS and to relate them with biological properties as the relative strength of the promoter. Our study shows how a model based on a coarse-grained description of a biomolecule can give valuable insight on its biological function.

Effective automated feature construction and selection for classification of biological sequences

BACKGROUND

Many open problems in bioinformatics involve elucidating underlying functional signals in biological sequences. DNA sequences, in particular, are characterized by rich architectures in which functional signals are increasingly found to combine local and distal interactions at the nucleotide level. Problems of interest include detection of regulatory regions, splice sites, exons, hypersensitive sites, and more. These problems naturally lend themselves to formulation as classification problems in machine learning. When classification is based on features extracted from the sequences under investigation, success is critically dependent on the chosen set of features.

METHODOLOGY

We present an algorithmic framework (EFFECT) for automated detection of functional signals in biological sequences. We focus here on classification problems involving DNA sequences which state-of-the-art work in machine learning shows to be challenging and involve complex combinations of local and distal features. EFFECT uses a two-stage process to first construct a set of candidate sequence-based features and then select a most effective subset for the classification task at hand. Both stages make heavy use of evolutionary algorithms to efficiently guide the search towards informative features capable of discriminating between sequences that contain a particular functional signal and those that do not.

RESULTS

To demonstrate its generality, EFFECT is applied to three separate problems of importance in DNA research: the recognition of hypersensitive sites, splice sites, and ALU sites. Comparisons with state-of-the-art algorithms show that the framework is both general and powerful. In addition, a detailed analysis of the constructed features shows that they contain valuable biological information about DNA architecture, allowing biologists and other researchers to directly inspect the features and potentially use the insights obtained to assist wet-laboratory studies on retainment or modification of a specific signal. Code, documentation, and all data for the applications presented here are provided for the community at http://www.cs.gmu.edu/~ashehu/?q=OurTools.

Unix interfaces, Kleisli, bucandin structure, etc. -- the heroic beginning of bioinformatics in Singapore

Remarkably, Singapore as one of today's hotspots for bioinformatics and computational biology research appeared de novo out of pioneering efforts of engaged local individuals in the early 90-s that, supported with increasing public funds from 1996 on, morphed into the present vibrant research community. This article brings to mind the pioneers, their first successes and early institutional developments.

Identification of a secondary promoter of CASP8 and its related transcription factor PURα

Caspase-8 (CASP8) is an essential initiator of apoptosis and is associated with many diseases in humans including esophageal squamous cell carcinoma. CASP8 produces a variety of transcripts, which might perform distinct functions. However, the cis and trans transcriptional determinants that control CASP8 expression remain poorly defined. Using a series of luciferase reporter assays, we identified a novel secondary promoter of CASP8 within chr2: 202,122,236 to 202,123,227 and 25 kb downstream of the previously described CASP8 promoter. ENCODE ChIP-seq data for this novel promoter region revealed several epigenetic features, including high levels of histone H3 lysine 27 acetylation and lysine 4 methylation, as well as low levels of CpG island methylation. We developed a mass spectrometry based strategy to identify transcription factors that contribute to the function of the secondary promoter. We found that the transcription activator protein PURα is specifically involved in the transcriptional activation of the secondary promoter and may exert its function by forming a complex with E2F-1 and RNA polymerase II. PURα can bind to both DNA and RNA, and functions in the initiation of DNA replication, regulation of transcription. We observed that knockdown of PURα expression decreased the transcriptional activity of the secondary promoter and mRNA expression of CASP8 isoform G. Although the physiologic roles of this secondary promoter remain unclear, our data may help explain the complexity of CASP8 transcription and suggest that the various caspase 8 isoforms may have distinct regulations and functions.

Role of DNA sequence based structural features of promoters in transcription initiation and gene expression

Regulatory information for transcription initiation is present in a stretch of genomic DNA, called the promoter region that is located upstream of the transcription start site (TSS) of the gene. The promoter region interacts with different transcription factors and RNA polymerase to initiate transcription and contains short stretches of transcription factor binding sites (TFBSs), as well as structurally unique elements. Recent experimental and computational analyses of promoter sequences show that they often have non-B-DNA structural motifs, as well as some conserved structural properties, such as stability, bendability, nucleosome positioning preference and curvature, across a class of organisms. Here, we briefly describe these structural features, the differences observed in various organisms and their possible role in regulation of gene expression.

Unravelling the hidden DNA structural/physical code provides novel insights on promoter location

Although protein recognition of DNA motifs in promoter regions has been traditionally considered as a critical regulatory element in transcription, the location of promoters, and in particular transcription start sites (TSSs), still remains a challenge. Here we perform a comprehensive analysis of putative core promoter sequences relative to non-annotated predicted TSSs along the human genome, which were defined by distinct DNA physical properties implemented in our ProStar computational algorithm. A representative sampling of predicted regions was subjected to extensive experimental validation and analyses. Interestingly, the vast majority proved to be transcriptionally active despite the lack of specific sequence motifs, indicating that physical signaling is indeed able to detect promoter activity beyond conventional TSS prediction methods. Furthermore, highly active regions displayed typical chromatin features associated to promoters of housekeeping genes. Our results enable to redefine the promoter signatures and analyze the diversity, evolutionary conservation and dynamic regulation of human core promoters at large-scale. Moreover, the present study strongly supports the hypothesis of an ancient regulatory mechanism encoded by the intrinsic physical properties of the DNA that may contribute to the complexity of transcription regulation in the human genome.

A composite method based on formal grammar and DNA structural features in detecting human polymerase II promoter region

An important step in understanding gene regulation is to identify the promoter regions where the transcription factor binding takes place. Predicting a promoter region de novo has been a theoretical goal for many researchers for a long time. There exists a number of in silico methods to predict the promoter region de novo but most of these methods are still suffering from various shortcomings, a major one being the selection of appropriate features of promoter region distinguishing them from non-promoters. In this communication, we have proposed a new composite method that predicts promoter sequences based on the interrelationship between structural profiles of DNA and primary sequence elements of the promoter regions. We have shown that a Context Free Grammar (CFG) can formalize the relationships between different primary sequence features and by utilizing the CFG, we demonstrate that an efficient parser can be constructed for extracting these relationships from DNA sequences to distinguish the true promoter sequences from non-promoter sequences. Along with CFG, we have extracted the structural features of the promoter region to improve upon the efficiency of our prediction system. Extensive experiments performed on different datasets reveals that our method is effective in predicting promoter sequences on a genome-wide scale and performs satisfactorily as compared to other promoter prediction techniques.

Ensemble approach combining multiple methods improves human transcription start site prediction

Background

The computational prediction of transcription start sites is an important unsolved problem. Some recent progress has been made, but many promoters, particularly those not associated with CpG islands, are still difficult to locate using current methods. These methods use different features and training sets, along with a variety of machine learning techniques and result in different prediction sets.

Results

We demonstrate the heterogeneity of current prediction sets, and take advantage of this heterogeneity to construct a two-level classifier ('Profisi Ensemble') using predictions from 7 programs, along with 2 other data sources. Support vector machines using 'full' and 'reduced' data sets are combined in an either/or approach. We achieve a 14% increase in performance over the current state-of-the-art, as benchmarked by a third-party tool.

Conclusions

Supervised learning methods are a useful way to combine predictions from diverse sources.

High sensitivity TSS prediction: estimates of locations where TSS cannot occur

Background

Although transcription in mammalian genomes can initiate from various genomic positions (e.g., 3′UTR, coding exons, etc.), most locations on genomes are not prone to transcription initiation. It is of practical and theoretical interest to be able to estimate such collections of non-TSS locations (NTLs). The identification of large portions of NTLs can contribute to better focusing the search for TSS locations and thus contribute to promoter and gene finding. It can help in the assessment of 5′ completeness of expressed sequences, contribute to more successful experimental designs, as well as more accurate gene annotation.

Methodology

Using comprehensive collections of Cap Analysis of Gene Expression (CAGE) and other transcript data from mouse and human genomes, we developed a methodology that allows us, by performing computational TSS prediction with very high sensitivity, to annotate, with a high accuracy in a strand specific manner, locations of mammalian genomes that are highly unlikely to harbor transcription start sites (TSSs). The properties of the immediate genomic neighborhood of 98,682 accurately determined mouse and 113,814 human TSSs are used to determine features that distinguish genomic transcription initiation locations from those that are not likely to initiate transcription. In our algorithm we utilize various constraining properties of features identified in the upstream and downstream regions around TSSs, as well as statistical analyses of these surrounding regions.

Conclusions

Our analysis of human chromosomes 4, 21 and 22 estimates ∼46%, ∼41% and ∼27% of these chromosomes, respectively, as being NTLs. This suggests that on average more than 40% of the human genome can be expected to be highly unlikely to initiate transcription. Our method represents the first one that utilizes high-sensitivity TSS prediction to identify, with high accuracy, large portions of mammalian genomes as NTLs. The server with our algorithm implemented is available at http://cbrc.kaust.edu.sa/ddm/.

SCS: signal, context, and structure features for genome-wide human promoter recognition

This paper integrates the signal, context, and structure features for genome-wide human promoter recognition, which is important in improving genome annotation and analyzing transcriptional regulation without experimental supports of ESTs, cDNAs, or mRNAs. First, CpG islands are salient biological signals associated with approximately 50 percent of mammalian promoters. Second, the genomic context of promoters may have biological significance, which is based on n-mers (sequences of n bases long) and their statistics estimated from training samples. Third, sequence-dependent DNA flexibility originates from DNA 3D structures and plays an important role in guiding transcription factors to the target site in promoters. Employing decision trees, we combine above signal, context, and structure features to build a hierarchical promoter recognition system called SCS. Experimental results on controlled data sets and the entire human genome demonstrate that SCS is significantly superior in terms of sensitivity and specificity as compared to other state-of-the-art methods. The SCS promoter recognition system is available online as supplemental materials for academic use and can be found on the Computer Society Digital Library at http://doi.ieeecomputersociety.org/10.1109/TCBB.2008.95.

High DNA melting temperature predicts transcription start site location in human and mouse

The accurate computational prediction of transcription start sites (TSS) in vertebrate genomes is a difficult problem. The physicochemical properties of DNA can be computed in various ways and a many combinations of DNA features have been tested in the past for use as predictors of transcription. We looked in detail at melting temperature, which measures the temperature, at which two strands of DNA separate, considering the cooperative nature of this process. We find that peaks in melting temperature correspond closely to experimentally determined transcription start sites in human and mouse chromosomes. Using melting temperature alone, and with simple thresholding, we can predict TSS with accuracy that is competitive with the most accurate state-of-the-art TSS prediction methods. Accuracy is measured using both experimentally and manually determined TSS. The method works especially well with CpG island containing promoters, but also works when CpG islands are absent. This result is clear evidence of the important role of the physical properties of DNA in the process of transcription. It also points to the importance for TSS prediction methods to include melting temperature as prior information.

Polymorphism in the flanking regions of the PbGP43 gene from the human pathogen Paracoccidioides brasiliensis: search for protein binding sequences and poly(A) cleavage sites

Background

Paracoccidioides brasiliensis is a thermo-dimorphic fungus that causes paracoccidiodomycosis (PCM). Glycoprotein gp43 is the fungal main diagnostic antigen, which can also protect against murine PCM and interact with extracellular matrix proteins. It is structurally related to glucanases, however not active, and whose expression varies considerably. We have presently studied polymorphisms in the PbGP43 flanking regions to help understand such variations.

Results

we tested the protein-binding capacity of oligonucleotides covering the PbGP43 proximal 5' flanking region, including overlap and mutated probes. We used electrophoretic mobility shift assays and found DNA binding regions between positions -134 to -103 and -255 to -215. Only mutation at -230, characteristic of P. brasiliensis phylogenetic species PS2, altered binding affinity. Next, we cloned and sequenced the 5' intergenic region up to position -2,047 from P. brasiliensis Pb339 and observed that it is composed of three tandem repetitive regions of about 500 bp preceded upstream by 442 bp. Correspondent PCR fragments of about 2,000 bp were found in eight out of fourteen isolates; in PS2 samples they were 1,500-bp long due to the absence of one repetitive region, as detected in Pb3. We also compared fifty-six PbGP43 3' UTR sequences from ten isolates and have not observed polymorphisms; however we detected two main poly(A) clusters (1,420 to 1,441 and 1,451 to 1,457) of multiple cleavage sites. In a single isolate we found one to seven sites.

Conclusions

We observed that the amount of PbGP43 transcripts accumulated in P. brasiliensis Pb339 grown in defined medium was about 1,000-fold higher than in Pb18 and 120-fold higher than in Pb3. We have described a series of features in the gene flanking regions and differences among isolates, including DNA-binding sequences, which might impact gene regulation. Little is known about regulatory sequences in thermo-dimorphic fungi. The peculiar structure of tandem repetitive fragments in the 5' intergenic region of PbGP43, their characteristic sequences, besides the presence of multiple poly(A) cleavage sites in the 3' UTR will certainly guide future studies.

Toward a gold standard for promoter prediction evaluation

Motivation: Promoter prediction is an important task in genome annotation projects, and during the past years many new promoter prediction programs (PPPs) have emerged. However, many of these programs are compared inadequately to other programs. In most cases, only a small portion of the genome is used to evaluate the program, which is not a realistic setting for whole genome annotation projects. In addition, a common evaluation design to properly compare PPPs is still lacking.

Results: We present a large-scale benchmarking study of 17 state-of-the-art PPPs. A multi-faceted evaluation strategy is proposed that can be used as a gold standard for promoter prediction evaluation, allowing authors of promoter prediction software to compare their method to existing methods in a proper way. This evaluation strategy is subsequently used to compare the chosen promoter predictors, and an in-depth analysis on predictive performance, promoter class specificity, overlap between predictors and positional bias of the predictions is conducted.

Availability: We provide the implementations of the four protocols, as well as the datasets required to perform the benchmarks to the academic community free of charge on request.

Contact:yves.vandepeer@psb.ugent.be

Supplementary information:Supplementary data are available at Bioinformatics online.

Towards accurate human promoter recognition: a review of currently used sequence features and classification methods

This review describes important advances that have been made during the past decade for genome-wide human promoter recognition. Interest in promoter recognition algorithms on a genome-wide scale is worldwide and touches on a number of practical systems that are important in analysis of gene regulation and in genome annotation without experimental support of ESTs, cDNAs or mRNAs. The main focus of this review is on feature extraction and model selection for accurate human promoter recognition, with descriptions of what they are, what has been accomplished, and what remains to be done.

Identifying regulatory elements in eukaryotic genomes

Proper development and functioning of an organism depends on precise spatial and temporal expression of all its genes. These coordinated expression-patterns are maintained primarily through the process of transcriptional regulation. Transcriptional regulation is mediated by proteins binding to regulatory elements on the DNA in a combinatorial manner, where particular combinations of transcription factor binding sites establish specific regulatory codes. In this review, we survey experimental and computational approaches geared towards the identification of proximal and distal gene regulatory elements in the genomes of complex eukaryotes. Available approaches that decipher the genetic structure and function of regulatory elements by exploiting various sources of information like gene expression data, chromatin structure, DNA-binding specificities of transcription factors, cooperativity of transcription factors, etc. are highlighted. We also discuss the relevance of regulatory elements in the context of human health through examples of mutations in some of these regions having serious implications in misregulation of genes and being strongly associated with human disorders.

Integrating sequence, evolution and functional genomics in regulatory genomics

With genome analysis expanding from the study of genes to the study of gene regulation, 'regulatory genomics' utilizes sequence information, evolution and functional genomics measurements to unravel how regulatory information is encoded in the genome.

Core promoters are predicted by their distinct physicochemical properties in the genome of Plasmodium falciparum

A method is presented to computationally identify core promoters in the Plasmodium falciparum genome using only DNA physicochemical properties.

Little is known about the structure and distinguishing features of core promoters in Plasmodium falciparum. In this work, we describe the first method to computationally identify core promoters in this AT-rich genome. This prediction algorithm uses solely DNA physicochemical properties as descriptors. Our results add to a growing body of evidence that a physicochemical code for eukaryotic genomes plays a crucial role in core promoter recognition.

High-resolution human core-promoter prediction with CoreBoost_HM

Correctly locating the gene transcription start site and the core-promoter is important for understanding transcriptional regulation mechanism. Here we have integrated specific genome-wide histone modification and DNA sequence features together to predict RNA polymerase II core-promoters in the human genome. Our new predictor CoreBoost_HM outperforms existing promoter prediction algorithms by providing significantly higher sensitivity and specificity at high resolution. We demonstrated that even though the histone modification data used in this study are from a specific cell type (CD4+ T-cell), our method can be used to identify both active and repressed promoters. We have applied it to search the upstream regions of microRNA genes, and show that CoreBoost_HM can accurately identify the known promoters of the intergenic microRNAs. We also identified a few intronic microRNAs that may have their own promoters. This result suggests that our new method can help to identify and characterize the core-promoters of both coding and noncoding genes.

How epigenomics contributes to the understanding of gene regulation in Toxoplasma gondii

How apicomplexan parasites regulate their gene expression is poorly understood. The complex life cycle of these parasites implies tight control of gene expression to orchestrate the appropriate expression pattern at the right moment. Recently, several studies have demonstrated the role of epigenetic mechanisms for control of coordinated expression of genes. In this review, we discuss the contribution of epigenomics to the understanding of gene regulation in Toxoplasma gondii. Studying the distribution of modified histones on the genome links chromatin modifications to gene expression or gene repression. In particular, coincident trimethylated lysine 4 on histone H3 (H3K4me3), acetylated lysine 9 on histone H3 (H3K9ac), and acetylated histone H4 (H4ac) mark promoters of actively transcribed genes. However, the presence of these modified histones at some non-expressed genes and other histone modifications at only a subset of active promoters implies the presence of other layers of regulation of chromatin structure in T. gondii. Epigenomics analysis provides a powerful tool to characterize the activation state of genomic loci of T. gondii and possibly of other Apicomplexa including Plasmodium or Cryptosporidium. Further, integration of epigenetic data with expression data and other genome-wide datasets facilitates refinement of genome annotation based upon experimental data.

ProSOM: core promoter prediction based on unsupervised clustering of DNA physical profiles

MOTIVATION

More and more genomes are being sequenced, and to keep up with the pace of sequencing projects, automated annotation techniques are required. One of the most challenging problems in genome annotation is the identification of the core promoter. Because the identification of the transcription initiation region is such a challenging problem, it is not yet a common practice to integrate transcription start site prediction in genome annotation projects. Nevertheless, better core promoter prediction can improve genome annotation and can be used to guide experimental work.

RESULTS

Comparing the average structural profile based on base stacking energy of transcribed, promoter and intergenic sequences demonstrates that the core promoter has unique features that cannot be found in other sequences. We show that unsupervised clustering by using self-organizing maps can clearly distinguish between the structural profiles of promoter sequences and other genomic sequences. An implementation of this promoter prediction program, called ProSOM, is available and has been compared with the state-of-the-art. We propose an objective, accurate and biologically sound validation scheme for core promoter predictors. ProSOM performs at least as well as the software currently available, but our technique is more balanced in terms of the number of predicted sites and the number of false predictions, resulting in a better all-round performance. Additional tests on the ENCODE regions of the human genome show that 98% of all predictions made by ProSOM can be associated with transcriptionally active regions, which demonstrates the high precision.

AVAILABILITY

Predictions for the human genome, the validation datasets and the program (ProSOM) are available upon request.

Profiling the thermodynamic softness of adenoviral promoters

We showed previously that anharmonic DNA dynamical features correlate with transcriptional activity in selected viral promoters, and hypothesized that areas of DNA softness may represent loci of functional significance. The nine known promoters from human adenovirus type 5 were analyzed for inherent DNA softness using the Peyrard-Bishop-Dauxois model and a statistical mechanics approach, using a transfer integral operator. We found a loosely defined pattern of softness peaks distributed both upstream and downstream of the transcriptional start sites, and that early transcriptional regions tended to be softer than late promoter regions. When reported transcription factor binding sites were superimposed on our calculated softness profiles, we observed a close correspondence in many cases, which suggests that DNA duplex breathing dynamics may play a role in protein recognition of specific nucleotide sequences and protein-DNA binding. These results suggest that genetic information is stored not only in explicit codon sequences, but also may be encoded into local dynamic and structural features, and that it may be possible to access this obscured information using DNA dynamics calculations.

Determining promoter location based on DNA structure first-principles calculations

A new method is presented which predicts promoter regions based on atomistic molecular dynamics simulations of small oligonucleotides, without requiring information on sequence conservation or features.

A new method for the prediction of promoter regions based on atomic molecular dynamics simulations of small oligonucleotides has been developed. The method works independently of gene structure conservation and orthology and of the presence of detectable sequence features. Results obtained with our method confirm the existence of a hidden physical code that modulates genome expression.

Generic eukaryotic core promoter prediction using structural features of DNA

Despite many recent efforts, in silico identification of promoter regions is still in its infancy. However, the accurate identification and delineation of promoter regions is important for several reasons, such as improving genome annotation and devising experiments to study and understand transcriptional regulation. Current methods to identify the core region of promoters require large amounts of high-quality training data and often behave like black box models that output predictions that are difficult to interpret. Here, we present a novel approach for predicting promoters in whole-genome sequences by using large-scale structural properties of DNA. Our technique requires no training, is applicable to many eukaryotic genomes, and performs extremely well in comparison with the best available promoter prediction programs. Moreover, it is fast, simple in design, and has no size constraints, and the results are easily interpretable. We compared our approach with 14 current state-of-the-art implementations using human gene and transcription start site data and analyzed the ENCODE region in more detail. We also validated our method on 12 additional eukaryotic genomes, including vertebrates, invertebrates, plants, fungi, and protists.

A code for transcription initiation in mammalian genomes

Genome-wide detection of transcription start sites (TSSs) has revealed that RNA Polymerase II transcription initiates at millions of positions in mammalian genomes. Most core promoters do not have a single TSS, but an array of closely located TSSs with different rates of initiation. As a rule, genes have more than one such core promoter; however, defining the boundaries between core promoters is not trivial. These discoveries prompt a re-evaluation of our models for transcription initiation. We describe a new framework for understanding the organization of transcription initiation. We show that initiation events are clustered on the chromosomes at multiple scales-clusters within clusters-indicating multiple regulatory processes. Within the smallest of such clusters, which can be interpreted as core promoters, the local DNA sequence predicts the relative transcription start usage of each nucleotide with a remarkable 91% accuracy, implying the existence of a DNA code that determines TSS selection. Conversely, the total expression strength of such clusters is only partially determined by the local DNA sequence. Thus, the overall control of transcription can be understood as a combination of large- and small-scale effects; the selection of transcription start sites is largely governed by the local DNA sequence, whereas the transcriptional activity of a locus is regulated at a different level; it is affected by distal features or events such as enhancers and chromatin remodeling.

Accurate splice site prediction using support vector machines

BACKGROUND

For splice site recognition, one has to solve two classification problems: discriminating true from decoy splice sites for both acceptor and donor sites. Gene finding systems typically rely on Markov Chains to solve these tasks.

RESULTS

In this work we consider Support Vector Machines for splice site recognition. We employ the so-called weighted degree kernel which turns out well suited for this task, as we will illustrate in several experiments where we compare its prediction accuracy with that of recently proposed systems. We apply our method to the genome-wide recognition of splice sites in Caenorhabditis elegans, Drosophila melanogaster, Arabidopsis thaliana, Danio rerio, and Homo sapiens. Our performance estimates indicate that splice sites can be recognized very accurately in these genomes and that our method outperforms many other methods including Markov Chains, GeneSplicer and SpliceMachine. We provide genome-wide predictions of splice sites and a stand-alone prediction tool ready to be used for incorporation in a gene finder.

AVAILABILITY

Data, splits, additional information on the model selection, the whole genome predictions, as well as the stand-alone prediction tool are available for download at http://www.fml.mpg.de/raetsch/projects/splice.

MetaProm: a neural network based meta-predictor for alternative human promoter prediction

BACKGROUND

De novo eukaryotic promoter prediction is important for discovering novel genes and understanding gene regulation. In spite of the great advances made in the past decade, recent studies revealed that the overall performances of the current promoter prediction programs (PPPs) are still poor, and predictions made by individual PPPs do not overlap each other. Furthermore, most PPPs are trained and tested on the most-upstream promoters; their performances on alternative promoters have not been assessed.

RESULTS

In this paper, we evaluate the performances of current major promoter prediction programs (i.e., PSPA, FirstEF, McPromoter, DragonGSF, DragonPF, and FProm) using 42,536 distinct human gene promoters on a genome-wide scale, and with emphasis on alternative promoters. We describe an artificial neural network (ANN) based meta-predictor program that integrates predictions from the current PPPs and the predicted promoters' relation to CpG islands. Our specific analysis of recently discovered alternative promoters reveals that although only 41% of the 3' most promoters overlap a CpG island, 74% of 5' most promoters overlap a CpG island.

CONCLUSION

Our assessment of six PPPs on 1.06 x 109 bps of human genome sequence reveals the specific strengths and weaknesses of individual PPPs. Our meta-predictor outperforms any individual PPP in sensitivity and specificity. Furthermore, we discovered that the 5' alternative promoters are more likely to be associated with a CpG island.

Boosting with stumps for predicting transcription start sites

CoreBoost applies a boosting technique to select important features for predicting core promoters with diverse patterns.

Promoter prediction is a difficult but important problem in gene finding, and it is critical for elucidating the regulation of gene expression. We introduce a new promoter prediction program, CoreBoost, which applies a boosting technique with stumps to select important small-scale as well as large-scale features. CoreBoost improves greatly on locating transcription start sites. We also demonstrate that by further utilizing some tissue-specific information, better accuracy can be achieved.

Identification and analysis of functional elements in 1% of the human genome by the ENCODE pilot project

We report the generation and analysis of functional data from multiple, diverse experiments performed on a targeted 1% of the human genome as part of the pilot phase of the ENCODE Project. These data have been further integrated and augmented by a number of evolutionary and computational analyses. Together, our results advance the collective knowledge about human genome function in several major areas. First, our studies provide convincing evidence that the genome is pervasively transcribed, such that the majority of its bases can be found in primary transcripts, including non-protein-coding transcripts, and those that extensively overlap one another. Second, systematic examination of transcriptional regulation has yielded new understanding about transcription start sites, including their relationship to specific regulatory sequences and features of chromatin accessibility and histone modification. Third, a more sophisticated view of chromatin structure has emerged, including its inter-relationship with DNA replication and transcriptional regulation. Finally, integration of these new sources of information, in particular with respect to mammalian evolution based on inter- and intra-species sequence comparisons, has yielded new mechanistic and evolutionary insights concerning the functional landscape of the human genome. Together, these studies are defining a path for pursuit of a more comprehensive characterization of human genome function.

The human genomic melting map

In a living cell, the antiparallel double-stranded helix of DNA is a dynamically changing structure. The structure relates to interactions between and within the DNA strands, and the array of other macromolecules that constitutes functional chromatin. It is only through its changing conformations that DNA can organize and structure a large number of cellular functions. In particular, DNA must locally uncoil, or melt, and become single-stranded for DNA replication, repair, recombination, and transcription to occur. It has previously been shown that this melting occurs cooperatively, whereby several base pairs act in concert to generate melting bubbles, and in this way constitute a domain that behaves as a unit with respect to local DNA single-strandedness. We have applied a melting map calculation to the complete human genome, which provides information about the propensities of forming local bubbles determined from the whole sequence, and present a first report on its basic features, the extent of cooperativity, and correlations to various physical and biological features of the human genome. Globally, the melting map covaries very strongly with GC content. Most importantly, however, cooperativity of DNA denaturation causes this correlation to be weaker at resolutions fewer than 500 bps. This is also the resolution level at which most structural and biological processes occur, signifying the importance of the informational content inherent in the genomic melting map. The human DNA melting map may be further explored at http://meltmap.uio.no.

Identification of plant promoter constituents by analysis of local distribution of short sequences

Background

Plant promoter architecture is important for understanding regulation and evolution of the promoters, but our current knowledge about plant promoter structure, especially with respect to the core promoter, is insufficient. Several promoter elements including TATA box, and several types of transcriptional regulatory elements have been found to show local distribution within promoters, and this feature has been successfully utilized for extraction of promoter constituents from human genome.

Results

LDSS (Local Distribution of Short Sequences) profiles of short sequences along the plant promoter have been analyzed in silico, and hundreds of hexamer and octamer sequences have been identified as having localized distributions within promoters of Arabidopsis thaliana and rice. Based on their localization patterns, the identified sequences could be classified into three groups, pyrimidine patch (Y Patch), TATA box, and REG (Regulatory Element Group). Sequences of the TATA box group are consistent with the ones reported in previous studies. The REG group includes more than 200 sequences, and half of them correspond to known cis-elements. The other REG subgroups, together with about a hundred uncategorized sequences, are suggested to be novel cis-regulatory elements. Comparison of LDSS-positive sequences between Arabidopsis and rice has revealed moderate conservation of elements and common promoter architecture. In addition, a dimer motif named the YR Rule (C/T A/G) has been identified at the transcription start site (-1/+1). This rule also fits both Arabidopsis and rice promoters.

Conclusion

LDSS was successfully applied to plant genomes and hundreds of putative promoter elements have been extracted as LDSS-positive octamers. Identified promoter architecture of monocot and dicot are well conserved, but there are moderate variations in the utilized sequences.

EGASP: the human ENCODE Genome Annotation Assessment Project

BACKGROUND

We present the results of EGASP, a community experiment to assess the state-of-the-art in genome annotation within the ENCODE regions, which span 1% of the human genome sequence. The experiment had two major goals: the assessment of the accuracy of computational methods to predict protein coding genes; and the overall assessment of the completeness of the current human genome annotations as represented in the ENCODE regions. For the computational prediction assessment, eighteen groups contributed gene predictions. We evaluated these submissions against each other based on a 'reference set' of annotations generated as part of the GENCODE project. These annotations were not available to the prediction groups prior to the submission deadline, so that their predictions were blind and an external advisory committee could perform a fair assessment.

RESULTS

The best methods had at least one gene transcript correctly predicted for close to 70% of the annotated genes. Nevertheless, the multiple transcript accuracy, taking into account alternative splicing, reached only approximately 40% to 50% accuracy. At the coding nucleotide level, the best programs reached an accuracy of 90% in both sensitivity and specificity. Programs relying on mRNA and protein sequences were the most accurate in reproducing the manually curated annotations. Experimental validation shows that only a very small percentage (3.2%) of the selected 221 computationally predicted exons outside of the existing annotation could be verified.

CONCLUSION

This is the first such experiment in human DNA, and we have followed the standards established in a similar experiment, GASP1, in Drosophila melanogaster. We believe the results presented here contribute to the value of ongoing large-scale annotation projects and should guide further experimental methods when being scaled up to the entire human genome sequence.