BiRen: predicting enhancers with a deep-learning-based model using the DNA sequence alone

Yang, Bite; Liu, Feng; Ren, Chao; Ouyang, Zhangyi; Xie, Ziwei; Bo, Xiaochen; Shu, Wenjie

doi:10.1093/bioinformatics/btx105

Cited by 130 publications

(104 citation statements)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Because the majority of GWAS SNPs are located within intergenic and intragenic enhancers (28,30,38) we tested whether GWAS disease risk located within the ketamine glutamate receptor and ketamine neuroplasticity sub-networks were likely to be located within chromatin loops in neural cells (SK-N-SH, H1, A735 cell lines and postmortem brain). If the GWAS SNPs contained in the sub-networks were predicted to be causal in the appropriate human surrogate tissue types, either neural (SK-N-SH, H1) or astrocyte (A735) cells but not in a liver cell line (HepG2) or in a white blood cell using multiple machine learning algorithms (39)(40)(41)(42)(43)(44)(45), they were used to probe public Hi-C datasets using a bioanalytic method developed in our laboratory (46) (Materials and Methods).…”

Section: Gwas Disease Risk Snps and Hi-c Loops Discriminate 2 Ketaminmentioning

confidence: 99%

“…These included both SNP signals significantly associated with disease risk and those associated with ketamine antidepressant response efficacy and dissociation, as well as other ketamine network and sub-network-associated SNPs from the published literature and clinical trials (8,37). The SNP filter judged candidate ketamine network SNPs for putative causality using multiple, redundant machine learning algorithms (39)(40)(41)(42)(43)(44)(45). The 65,535 SNPs contained in the EBI-NHGRI GWAS catalog (37) were evaluated for the 110-putative ketamineresponse genes and regulatory RNAs contained in our sub-networks.…”

Section: Selection Of Candidate Ketamine Response Snpsmentioning

confidence: 99%

See 1 more Smart Citation

Ketamine’s pharmacogenomic network in human brain contains sub-networks associated with glutamate neurotransmission and with neuroplasticity

Higgins

Handelman

Allyn-Feurer

et al. 2020

Preprint

View full text Add to dashboard Cite

The pharmacogenomic network responsible for the rapid antidepressant action of ketamine and concomitant adverse events in patients has been poorly defined. Integrative, multiscale biological data analytics helps explain ketamine's action. Using a validated computational pipeline, candidate ketamine-response genes and regulatory RNAs from published literature, binding affinity studies, and single nucleotide polymorphisms (SNPs) from genomewide association studies (GWAS), we identified 108 SNPs associated with 110 genes and regulatory RNAs. All of these SNPs are classified as enhancers, and additional chromatin interaction mapping in human neural cell lines and tissue shows enhancer-promoter interactions involving other network members. Pathway analysis and gene set optimization identified three composite sub-networks within the broader ketamine pharmacogenomic network. Expression patterns of ketamine network genes within the postmortem human brain are concordant with ketamine neurocircuitry based on the results of 24 published functional neuroimaging studies. The ketamine pharmacogenomic network is enriched in forebrain regions known to be rapidly activated by ketamine, including cingulate cortex and frontal cortex, and is significantly regulated by ketamine (p=6.26E-33; Fisher's exact test). The ketamine pharmacogenomic network can be partitioned into distinct enhancer sub-networks associated with: (1) glutamate neurotransmission, chromatin remodeling, smoking behavior, schizophrenia, pain, nausea, vomiting, and post-operative delirium; (2) neuroplasticity, depression, and alcohol consumption; and (3) pharmacokinetics. The component sub-networks explain the diverse action mechanisms of ketamine and its analogs.These results may be useful for optimizing pharmacotherapy in patients diagnosed with depression, pain or related stress disorders. GRIA4 (12), and many other known (8,13) and unknown pharmacodynamic targets within human brain. Ketamine and its metabolites strongly induce the expression of the CYP2B6 gene in human brain, which encodes a drug metabolizing enzyme that contributes to first-and second-pass metabolism of the drug and its metabolites (14). Recent genomewide association studies (GWAS) in humans demonstrate association of ketamine response and adverse events with enhancers of genes and long non-coding RNAs (lncRNAs) related to the roundabout guidance receptor 2 (ROBO2) gene, whose product binds members of the slit guidance ligand family (SLIT1, SLIT2) that are involved in dendrite guidance and synaptic plasticity (15,16). Like phencyclidine, a structurally related compound, ketamine induces acute dissociation, with both drugs exhibiting species-specific differences in response. In sum, the central nervous system (CNS) pathway(s) responsible for the rapid antidepressant effects of ketamine and its enantiomers in patients diagnosed with treatment-resistant depression (TRD) remain poorly defined, including emergence of on-and off-target effects and individual differences in response and adverse eve...

show abstract

Section: Gwas Disease Risk Snps and Hi-c Loops Discriminate 2 Ketaminmentioning

confidence: 99%

Section: Selection Of Candidate Ketamine Response Snpsmentioning

confidence: 99%

Ketamine’s pharmacogenomic network in human brain contains sub-networks associated with glutamate neurotransmission and with neuroplasticity

Higgins

Handelman

Allyn-Feurer

et al. 2020

Preprint

View full text Add to dashboard Cite

show abstract

“…In recent years, deep neural networks (DNNs) have achieved state-of-art prediction =accuracies for many tasks in regulatory genomics, such as predicting splicing activity (Leung et al, 2014; Xiong et al, 2015), specificities of DNA- and RNA-binding proteins (Alipanahi et al, 2015), transcription factor binding sites (TFBS) (Quang and Xie, 2019, 2015; Zhou and Troyanskaya, 2015), epigenetic marks (Kelley et al, 2016; Quang and Xie, 2015; Zhou and Troyanskaya, 2015), enhancer activity (Min et al, 2016; Yang et al, 2017) and enhancer-promoter interactions (Singh and Yang, 2016). However, in spite of their superior performance, little biological knowledge or mechanistic understanding has been gained from DNN models.…”

Section: Introductionmentioning

confidence: 99%

Learning and Interpreting the Gene Regulatory Grammar in a Deep Learning Framework

Capra

Chen

2019

Preprint

View full text Add to dashboard Cite

10Deep neural networks (DNNs) have achieved state-of-the-art performance in identifying gene 11 regulatory sequences, but they have provided limited insight into the biology of regulatory 12 elements due to the difficulty of interpreting the complex features they learn. Several models of 13 how combinatorial binding of transcription factors, i.e. the regulatory grammar, drives enhancer 14 activity have been proposed, ranging from the flexible TF billboard model to the stringent 15 enhanceosome model. However, there is limited knowledge of the prevalence of these (or other) 16 sequence architectures across enhancers. Here we perform several hypothesis-driven analyses to 17 explore the ability of DNNs to learn the regulatory grammar of enhancers. We created a 18 synthetic dataset based on existing hypotheses about combinatorial transcription factor binding 19 site (TFBS) patterns, including homotypic clusters, heterotypic clusters, and enhanceosomes, 20 from real TF binding motifs from diverse TF families. We then trained deep residual neural 21 networks (ResNets) to model the sequences under a range of scenarios that reflect real-world 22 multi-label regulatory sequence prediction tasks. We developed a gradient-based unsupervised 23 clustering method to extract the patterns learned by the ResNet models. We demonstrated that 24 simulated regulatory grammars are best learned in the penultimate layer of the ResNets, and the 25 proposed method can accurately retrieve the regulatory grammar even when there is 26 heterogeneity in the enhancer categories and a large fraction of TFBS outside of the regulatory 27 grammar. However, we also identify common scenarios where ResNets fail to learn simulated 28 regulatory grammars. Our results provide a framework for interpreting the regulatory rules 29 learned by ResNets, and they demonstrate that the ability and efficiency of ResNets in learning 30 the regulatory grammar highly depends on the nature of the prediction task. 31

show abstract

“…Hybrid approahes combining Convolutional and Recurrent Neural Netowrks were used to predict enchansers in DNA. The Biren [40] method took solely the DNA sequence and did the rest of the processing on its own. The concept of DL has been used in several other researches and is still being used till date due to its efficiency and novelty.…”

Section: Introductionmentioning

confidence: 99%

DeepDBP: Deep Neural Networks for Identification of DNA-binding Proteins

Shadab

Khan

Neezi

et al. 2019

Preprint

View full text Add to dashboard Cite

DNA-Binding proteins (DBP) are associated with many cellular level functions which includes but not limited to body's defense mechanism and oxygen transportation.They bind DNAs and interact with them. In the past DBPs were identified using experimental lab based methods. However, in the recent years researchers are using supervised learning to identify DBPs solely from protein sequences. In this paper, we apply deep learning methods to identify DBPs. We have proposed two different deep learning based methods for identifying DBPs: DeepDBP-ANN and DeepDBP-CNN. DeepDBP-ANN uses a generated set of features trained on traditional neural network and DeepDBP-CNN uses a pre-learned embedding and Convolutional Neural Network. Both of our proposed methods were able to produce state-of-the-art results when tested on standard benchmark datasets.DeepDBP-ANN had a train accuracy of 99.02% and test accuracy of 82.80%.And DeepDBP-CNN though had train accuracy of 94.32%, it excelled at identifying test instances with 84.31% accuracy. All methods are available codes and methods are available for use at: https://github.com/antorkhan/DNABinding.

show abstract

BiRen: predicting enhancers with a deep-learning-based model using the DNA sequence alone

Abstract: Supplementary data are available at Bioinformatics online.

Cited by 130 publications

References 34 publications

Ketamine’s pharmacogenomic network in human brain contains sub-networks associated with glutamate neurotransmission and with neuroplasticity

Ketamine’s pharmacogenomic network in human brain contains sub-networks associated with glutamate neurotransmission and with neuroplasticity

Learning and Interpreting the Gene Regulatory Grammar in a Deep Learning Framework

DeepDBP: Deep Neural Networks for Identification of DNA-binding Proteins

Contact Info

Product

Resources

About