DeepCAPE: A Deep Convolutional Neural Network for the Accurate Prediction of Enhancers

The establishment of a landscape of enhancers across human cells is crucial to deciphering the mechanism of gene regulation, cell differentiation, and disease development. High-throughput experimental approaches, which contain successfully reported enhancers in typical cell lines, are still too costly and time-consuming to perform systematic identification of enhancers specific to different cell lines. Existing computational methods, capable of predicting regulatory elements purely relying on DNA sequences, lack the power of cell line-specific screening. Recent studies have suggested that chromatin accessibility of a DNA segment is closely related to its potential function in regulation, and thus may provide useful information in identifying regulatory elements. Motivated by the aforementioned understanding, we integrate DNA sequences and chromatin accessibility data to accurately predict enhancers in a cell line-specific manner. We proposed DeepCAPE, a deep convolutional neural network to predict enhancers via the integration of DNA sequences and DNase-seq data. Benefitting from the well-designed feature extraction mechanism and skip connection strategy, our model not only consistently outperforms existing methods in the imbalanced classification of cell line-specific enhancers against background sequences, but also has the ability to self-adapt to different sizes of datasets. Besides, with the adoption of auto-encoder, our model is capable of making cross-cell line predictions. We further visualize kernels of the first convolutional layer and show the match of identified sequence signatures and known motifs. We finally demonstrate the potential ability of our model to explain functional implications of putative disease-associated genetic variants and discriminate disease-related enhancers. The source code and detailed tutorial of DeepCAPE are freely available at https://github.com/ShengquanChen/DeepCAPE.     

The chromatin fingerprint of gene enhancer elements

Different cell types within a single organism are generally distinguished by strikingly different patterns of gene expression, which are dynamic throughout development and adult life. Distal enhancer elements are key drivers of spatiotemporal specificity in gene regulation. Often located tens of kilobases from their target promoters and functioning in an orientation-independent manner, the identification of bona fide enhancers has proved a formidable challenge. With the development of ChIP-seq, global cataloging of putative enhancers has become feasible. Here, we review the current understanding of the chromatin landscape at enhancers and how these chromatin features enable robust identification of tissue-specific enhancers.     

Long ncRNA expression associates with tissue-specific enhancers

Long non-coding RNAs (ncRNA) have recently been demonstrated to be expressed from a subset of enhancers and to be required for the distant regulation of gene expression. Several approaches to predict enhancers have been developed based on various chromatin marks and occupancy of enhancer-binding proteins. Despite the rapid advances in the field, no consensus how to define tissue specific enhancers yet exists. Here, we identify 2,695 long ncRNAs annotated by ENCODE (corresponding to 28% of all ENCODE annotated long ncRNAs) that overlap tissue-specific enhancers. We use a recently developed algorithm to predict tissue-specific enhancers, PreSTIGE, that is based on the H3K4me1 mark and tissue specific expression of mRNAs. The expression of the long ncRNAs overlapping enhancers is significantly higher when the enhancer is predicted as active in a specific cell line, suggesting a general interdependency of active enhancers and expression of long ncRNAs. This dependency is not identified using previous enhancer prediction algorithms that do not account for expression of their downstream targets. The predicted enhancers that overlap annotated long ncRNAs generally have a lower ratio of H3K4me1 to H3K4me3, suggesting that enhancers expressing long ncRNAs might be associated with specific epigenetic marks. In conclusion, we demonstrate the tissue-specific predictive power of PreSTIGE and provide evidence for thousands of long ncRNAs that are expressed from active tissue-specific enhancers, suggesting a particularly important functional relationship between long ncRNAs and enhancer activity in determining tissue-specific gene expression.     

Genome-wide enhancer prediction from epigenetic signatures using genetic algorithm-optimized support vector machines

The chemical modification of histones at specific DNA regulatory elements is linked to the activation, inactivation and poising of genes. A number of tools exist to predict enhancers from chromatin modification maps, but their practical application is limited because they either (i) consider a smaller number of marks than those necessary to define the various enhancer classes or (ii) work with an excessive number of marks, which is experimentally unviable. We have developed a method for chromatin state detection using support vector machines in combination with genetic algorithm optimization, called ChromaGenSVM. ChromaGenSVM selects optimum combinations of specific histone epigenetic marks to predict enhancers. In an independent test, ChromaGenSVM recovered 88% of the experimentally supported enhancers in the pilot ENCODE region of interferon gamma-treated HeLa cells. Furthermore, ChromaGenSVM successfully combined the profiles of only five distinct methylation and acetylation marks from ChIP-seq libraries done in human CD4⁺ T cells to predict ∼21 000 experimentally supported enhancers within 1.0 kb regions and with a precision of ∼90%, thereby improving previous predictions on the same dataset by 21%. The combined results indicate that ChromaGenSVM comfortably outperforms previously published methods and that enhancers are best predicted by specific combinations of histone methylation and acetylation marks.     

Predicting enhancer-promoter interaction from genomic sequence with deep neural networks

Background: In the human genome, distal enhancers are involved in regulating target genes through proximal promoters by forming enhancer-promoter interactions. Although recently developed high-throughput experimental approaches have allowed us to recognize potential enhancer-promoter interactions genome-wide, it is still largely unclear to what extent the sequence-level information encoded in our genome help guide such interactions. Methods: Here we report a new computational method (named “SPEID”) using deep learning models to predict enhancer-promoter interactions based on sequence-based features only, when the locations of putative enhancers and promoters in a particular cell type are given. Results: Our results across six different cell types demonstrate that SPEID is effective in predicting enhancer-promoter interactions as compared to state-of-the-art methods that only use information from a single cell type. As a proof-of-principle, we also applied SPEID to identify somatic non-coding mutations in melanoma samples that may have reduced enhancer-promoter interactions in tumor genomes. Conclusions: This work demonstrates that deep learning models can help reveal that sequence-based features alone are sufficient to reliably predict enhancer-promoter interactions genome-wide.