Predicting enhancers in mammalian genomes using supervised hidden Markov models

Abstract

Background Eukaryotic gene regulation is a complex process comprising the dynamic interaction of enhancers and promoters in order to activate gene expression. In recent years, research in regulatory genomics has contributed to a better understanding of the characteristics of promoter elements and for most sequenced model organism genomes there exist comprehensive and reliable promoter annotations. For enhancers, however, a reliable description of their characteristics and location has so far proven to be elusive. With the development of high-throughput methods such as ChIP-seq, large amounts of data about epigenetic conditions have become available, and many existing methods use the information on chromatin accessibility or histone modifications to train classifiers in order to segment the genome into functional groups such as enhancers and promoters. However, these methods often do not consider prior biological knowledge about enhancers such as their diverse lengths or molecular structure. Results We developed enhancer HMM (eHMM), a supervised hidden Markov model designed to learn the molecular structure of promoters and enhancers. Both consist of a central stretch of accessible DNA flanked by nucleosomes with distinct histone modification patterns. We evaluated the performance of eHMM within and across cell types and developmental stages and found that eHMM successfully predicts enhancers with high precision and recall comparable to state-of-the-art methods, and consistently outperforms those in terms of accuracy and resolution. Conclusions eHMM predicts active enhancers based on data from chromatin accessibility assays and a minimal set of histone modification ChIP-seq experiments. In comparison to other ’black box’ methods its parameters are easy to interpret. eHMM can be used as a stand-alone tool for enhancer prediction without the need for additional training or a tuning of parameters. The high spatial precision of enhancer predictions gives valuable targets for potential knockout experiments or downstream analyses such as motif search.

Segmentation and genome annotation algorithms for identifying chromatin state and other genomic patterns

“…The analyst may downsample 1-bp resolution signal into bins (see “Spatial resolution”). This involves computing one of: ● average read count, ● reads per million mapped reads fold enrichment [ 47 ], ● total count of reads [ 19 , 48 , 49 ], or ● maximum count of reads of each bin [ 9 , 21 ]. Binning greatly decreases the computational cost of the SAGA algorithm and can improve the data’s statistical properties.…”

Fish-Ing for Enhancers in the Heart

“…However, HMM method predicted a higher number of TSS compared to profile-based approach, and its predictions were supported by other marks such as CAGE tags, P300, and DHSs. Furthermore, several popular HMM and their generalized form, dynamic Bayesian networks (DBN)-based mathematical models, have been developed for genome-wide enhancer prediction using either supervised (such as CHROMatin-based Integrated Approach (Chromia) [ 168 ] and enhancer-HMM [ 169 ]) or unsupervised (such as ChromHMM [ 170 ], Genostan [ 171 ], and Segway [ 172 ]) learning algorithms. The ENCODE project consortium [ 173 ] implemented an unsupervised machine learning method to annotate functionally relevant regions across 1640 genomics datasets in 147 distinct human cell-types, generated under this project.…”

Enhancers as potential targets for engineering salinity stress tolerance in crop plants

“…Recently, a few more computational methods have been developed with additional applications and/or advantages. For example, a supervised hidden Markov model (HMM) based method, enhancer HMM (eHMM), have been developed that can distinguish between enhancers and promoters too, despite a substantial overlap between their features (Zehnder et al 2019). In addition, methods for identification and classification of enhancers via dimension reduction technology and/or recurrent/ convolutional neural networks have been proposed (Li et al 2020a(Li et al , 2020b(Li et al , 2020cNguyen et al 2019).…”