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

Physical Peculiarity of Two Sites in Human Promoters: Universality and Diverse Usage in Gene Function

Since the discovery of physical peculiarities around transcription start sites (TSSs) and a site corresponding to the TATA box, research has revealed only the average features of these sites. Unsettled enigmas include the individual genes with these features and whether they relate to gene function. Herein, using 10 physical properties of DNA, including duplex DNA free energy, base stacking energy, protein-induced deformability, and stabilizing energy of Z-DNA, we clarified for the first time that approximately 97% of the promoters of 21,056 human protein-coding genes have distinctive physical properties around the TSS and/or position -27; of these, nearly 65% exhibited such properties at both sites. Furthermore, about 55% of the 21,056 genes had a minimum value of regional duplex DNA free energy within TSS-centered ±300 bp regions. Notably, distinctive physical properties within the promoters and free energies of the surrounding regions separated human protein-coding genes into five groups; each contained specific gene ontology (GO) terms. The group represented by immune response genes differed distinctly from the other four regarding the parameter of the free energies of the surrounding regions. A vital suggestion from this study is that physical-feature-based analyses of genomes may reveal new aspects of the organization and regulation of genes.

A successful hybrid deep learning model aiming at promoter identification

BACKGROUND

The zone adjacent to a transcription start site (TSS), namely, the promoter, is primarily involved in the process of DNA transcription initiation and regulation. As a result, proper promoter identification is critical for further understanding the mechanism of the networks controlling genomic regulation. A number of methodologies for the identification of promoters have been proposed. Nonetheless, due to the great heterogeneity existing in promoters, the results of these procedures are still unsatisfactory. In order to establish additional discriminative characteristics and properly recognize promoters, we developed the hybrid model for promoter identification (HMPI), a hybrid deep learning model that can characterize both the native sequences of promoters and the morphological outline of promoters at the same time. We developed the HMPI to combine a method called the PSFN (promoter sequence features network), which characterizes native promoter sequences and deduces sequence features, with a technique referred to as the DSPN (deep structural profiles network), which is specially structured to model the promoters in terms of their structural profile and to deduce their structural attributes.

RESULTS

The HMPI was applied to human, plant and Escherichia coli K-12 strain datasets, and the findings showed that the HMPI was successful at extracting the features of the promoter while greatly enhancing the promoter identification performance. In addition, after the improvements of synthetic sampling, transfer learning and label smoothing regularization, the improved HMPI models achieved good results in identifying subtypes of promoters on prokaryotic promoter datasets.

CONCLUSIONS

The results showed that the HMPI was successful at extracting the features of promoters while greatly enhancing the performance of identifying promoters on both eukaryotic and prokaryotic datasets, and the improved HMPI models are good at identifying subtypes of promoters on prokaryotic promoter datasets. The HMPI is additionally adaptable to different biological functional sequences, allowing for the addition of new features or models.

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.

DeepGSR: an optimized deep-learning structure for the recognition of genomic signals and regions

MOTIVATION

Recognition of different genomic signals and regions (GSRs) in DNA is crucial for understanding genome organization, gene regulation, and gene function, which in turn generate better genome and gene annotations. Although many methods have been developed to recognize GSRs, their pure computational identification remains challenging. Moreover, various GSRs usually require a specialized set of features for developing robust recognition models. Recently, deep-learning (DL) methods have been shown to generate more accurate prediction models than 'shallow' methods without the need to develop specialized features for the problems in question. Here, we explore the potential use of DL for the recognition of GSRs.

RESULTS

We developed DeepGSR, an optimized DL architecture for the prediction of different types of GSRs. The performance of the DeepGSR structure is evaluated on the recognition of polyadenylation signals (PAS) and translation initiation sites (TIS) of different organisms: human, mouse, bovine and fruit fly. The results show that DeepGSR outperformed the state-of-the-art methods, reducing the classification error rate of the PAS and TIS prediction in the human genome by up to 29% and 86%, respectively. Moreover, the cross-organisms and genome-wide analyses we performed, confirmed the robustness of DeepGSR and provided new insights into the conservation of examined GSRs across species.

AVAILABILITY AND IMPLEMENTATION

DeepGSR is implemented in Python using Keras API; it is available as open-source software and can be obtained at https://doi.org/10.5281/zenodo.1117159.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

DCDE: An Efficient Deep Convolutional Divergence Encoding Method for Human Promoter Recognition

Efficient human promoter feature extraction is still a major challenge in genome analysis as it can better understand human gene regulation and will be useful for experimental guidance. Although many machine learning algorithms have been developed for eukaryotic gene recognition, performance on promoters is unsatisfactory due to the diverse nature. To extract discriminative features from human promoters, an efficient deep convolutional divergence encoding method (DCDE) is proposed based on statistical divergence (SD) and convolutional neural network (CNN). SD can help optimize kmer feature extraction for human promoters. CNN can also be used to automatically extract features in gene analysis. In DCDE, we first perform informative kmers settlement to encode original gene sequences. A series of SD methods can optimize the most discriminative kmers distributions while maintaining important positional information. Then, CNN is utilized to extract lower dimensional deep features by secondary encoding. Finally, we construct a hybrid recognition architecture with multiple support vector machines and a bilayer decision method. It is flexible to add new features or new models and can be extended to identify other genomic functional elements. The extensive experiments demonstrate that DCDE is effective in promoter encoding and can significantly improve the performance of promoter recognition.

A Review of Computational Intelligence Methods for Eukaryotic Promoter Prediction

In past decades, prediction of genes in DNA sequences has attracted the attention of many researchers but due to its complex structure it is extremely intricate to correctly locate its position. A large number of regulatory regions are present in DNA that helps in transcription of a gene. Promoter is one such region and to find its location is a challenging problem. Various computational methods for promoter prediction have been developed over the past few years. This paper reviews these promoter prediction methods. Several difficulties and pitfalls encountered by these methods are also detailed, along with future research directions.

SD-MSAEs: Promoter recognition in human genome based on deep feature extraction

The prediction and recognition of promoter in human genome play an important role in DNA sequence analysis. Entropy, in Shannon sense, of information theory is a multiple utility in bioinformatic details analysis. The relative entropy estimator methods based on statistical divergence (SD) are used to extract meaningful features to distinguish different regions of DNA sequences. In this paper, we choose context feature and use a set of methods of SD to select the most effective n-mers distinguishing promoter regions from other DNA regions in human genome. Extracted from the total possible combinations of n-mers, we can get four sparse distributions based on promoter and non-promoters training samples. The informative n-mers are selected by optimizing the differentiating extents of these distributions. Specially, we combine the advantage of statistical divergence and multiple sparse auto-encoders (MSAEs) in deep learning to extract deep feature for promoter recognition. And then we apply multiple SVMs and a decision model to construct a human promoter recognition method called SD-MSAEs. Framework is flexible that it can integrate new feature extraction or new classification models freely. Experimental results show that our method has high sensitivity and specificity.

Rare k-mer DNA: Identification of sequence motifs and prediction of CpG island and promoter

Empirical analysis on k-mer DNA has been proven as an effective tool in finding unique patterns in DNA sequences which can lead to the discovery of potential sequence motifs. In an extensive study of empirical k-mer DNA on hundreds of organisms, the researchers found unique multi-modal k-mer spectra occur in the genomes of organisms from the tetrapod clade only which includes all mammals. The multi-modality is caused by the formation of the two lowest modes where k-mers under them are referred as the rare k-mers. The suppression of the two lowest modes (or the rare k-mers) can be attributed to the CG dinucleotide inclusions in them. Apart from that, the rare k-mers are selectively distributed in certain genomic features of CpG Island (CGI), promoter, 5' UTR, and exon. We correlated the rare k-mers with hundreds of annotated features using several bioinformatic tools, performed further intrinsic rare k-mer analyses within the correlated features, and modeled the elucidated rare k-mer clustering feature into a classifier to predict the correlated CGI and promoter features. Our correlation results show that rare k-mers are highly associated with several annotated features of CGI, promoter, 5' UTR, and open chromatin regions. Our intrinsic results show that rare k-mers have several unique topological, compositional, and clustering properties in CGI and promoter features. Finally, the performances of our RWC (rare-word clustering) method in predicting the CGI and promoter features are ranked among the top three, in eight of the CGI and promoter evaluations, among eight of the benchmarked datasets.

IN-MACA-MCC: Integrated Multiple Attractor Cellular Automata with Modified Clonal Classifier for Human Protein Coding and Promoter Prediction

Protein coding and promoter region predictions are very important challenges of bioinformatics (Attwood and Teresa, 2000). The identification of these regions plays a crucial role in understanding the genes. Many novel computational and mathematical methods are introduced as well as existing methods that are getting refined for predicting both of the regions separately; still there is a scope for improvement. We propose a classifier that is built with MACA (multiple attractor cellular automata) and MCC (modified clonal classifier) to predict both regions with a single classifier. The proposed classifier is trained and tested with Fickett and Tung (1992) datasets for protein coding region prediction for DNA sequences of lengths 54, 108, and 162. This classifier is trained and tested with MMCRI datasets for protein coding region prediction for DNA sequences of lengths 252 and 354. The proposed classifier is trained and tested with promoter sequences from DBTSS (Yamashita et al., 2006) dataset and nonpromoters from EID (Saxonov et al., 2000) and UTRdb (Pesole et al., 2002) datasets. The proposed model can predict both regions with an average accuracy of 90.5% for promoter and 89.6% for protein coding region predictions. The specificity and sensitivity values of promoter and protein coding region predictions are 0.89 and 0.92, respectively.

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

We comment on the flexibility profiles calculated by Zeng et al., and show that these profiles do not represent the local flexibility of the DNA molecule. If one takes into account the physics of elasticity, the averaged flexibility profile show an additional peak which is missed in the original calculation. We show that it is not possible to calculate the flexibility of a 6-mer using tetranucleotide elastic constants, the shortest sequence is a 7-mer. For 6-mers, dinucleotide or trinucleotide parameters are needed. We present calculations for dinucleotide flexibility parameters and show that the same additional peak is present for both 7-mers and 6-mers.

Multiconstrained gene clustering based on generalized projections

Background

Gene clustering for annotating gene functions is one of the fundamental issues in bioinformatics. The best clustering solution is often regularized by multiple constraints such as gene expressions, Gene Ontology (GO) annotations and gene network structures. How to integrate multiple pieces of constraints for an optimal clustering solution still remains an unsolved problem.

Results

We propose a novel multiconstrained gene clustering (MGC) method within the generalized projection onto convex sets (POCS) framework used widely in image reconstruction. Each constraint is formulated as a corresponding set. The generalized projector iteratively projects the clustering solution onto these sets in order to find a consistent solution included in the intersection set that satisfies all constraints. Compared with previous MGC methods, POCS can integrate multiple constraints from different nature without distorting the original constraints. To evaluate the clustering solution, we also propose a new performance measure referred to as Gene Log Likelihood (GLL) that considers genes having more than one function and hence in more than one cluster. Comparative experimental results show that our POCS-based gene clustering method outperforms current state-of-the-art MGC methods.

Conclusions

The POCS-based MGC method can successfully combine multiple constraints from different nature for gene clustering. Also, the proposed GLL is an effective performance measure for the soft clustering solutions.