Unsupervised pattern discovery in human chromatin structure through genomic segmentation

By Michael M. Hoffman, Orion J. Buske, Jie Wang, Zhiping Weng, J. Bilmes, William Stafford Noble. 2012

623 citations recorded by the source corpus.

We trained Segway, a dynamic Bayesian network method, simultaneously on chromatin data from multiple experiments, including positions of histone modifications, transcription-factor binding and open chromatin, all derived from a human chronic myeloid leukemia cell line. In an unsupervised fashion, we identified patterns associated with transcription start sites, gene ends, enhancers, transcriptional regulator CTCF-binding regions and repressed regions. Software and genome browser tracks are at http://noble.gs.washington.edu/proj/segway/.

Subfields: Molecular Biology

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

CADD: predicting the deleteriousness of variants throughout the human genome (2018, methodology): “Segway (16): replaced with Ensembl regulatory build (which is based on a Segway model)”
A general framework for estimating the relative pathogenicity of human genetic variants (2014, background): “To the 6 VEP input derived columns (chromosome, start, reference allele, alternative allele, variant type: SNV/INS/DEL, length) and 26 actual VEP output derived columns, we added 56 columns providing diverse annotations (e.g. mapability scores and segmental duplication annotation as distributed by U”
FactorNet: a deep learning framework for predicting cell type specific transcription factor binding from nucleotide-resolution sequential data (2017, methodology): “Some algorithms accomplish this by dividing the entire genome systematically into segments, and then assigning the resulting genome segments into “chromatin states” by applying machine learning methods such as Hidden Markov Models, Dynamic Bayesian Networks, or SelfOrganizing Maps [8, 9, 10].”
Genome-wide prediction of cis-regulatory regions using supervised deep learning methods (2016, methodology): “For instance, both ChromHMM [8] and Segway [9] segment the genome into sequence classes based on ENCODE project data [10], such as histone modiﬁcation ChIP-seq (chromatin immunoprecipitation followed by sequencing [11]) signals.”
An Integrated Encyclopedia of DNA Elements in the Human Genome (2012, background): “In the second approach, two methodologically distinct unbiased approaches (see ref 40,69 and M.M. Hoffman et al. , manuscript in preparation) converged on a concordant set of histone modification and chromatin-accessibility patterns that can be used to segment the genome in each of the Tier 1 and…”
The Ensembl Regulatory Build (2015, methodology): “This can be done with segmentation tools, such as Seg-way [29] or ChromHMM [30].”
Learning the human chromatin network from all ENCODE ChIP-seq data (2015, methodology): “We chose three (out of seven) Segway [20, 53] annotation labels that highlight distinct areas of the network (Figure 6C).”
Integrative annotation of chromatin elements from ENCODE data (2012, neutral): “For the second phase of ENCODE, two research groups within the consortium independently developed chromatin state annotation algorithms, ChromHMM (13,14) and Segway (15).”
Ensembl 2013 (2012, methodology): “In Ensembl release 65 (December 2011), we introduced the combined Segway (35) and ChromHMM (36) segmentation analyses developed for ENCODE (9), which classifies the genome into regions based on 12 specific assays to obtain a single-track summary of the functional architecture of the human genome.”
High-throughput functional testing of ENCODE segmentation predictions (2014, neutral): “These unsupervised learning algorithms, Segway (Hoffman et al. 2012) and ChromHMM (Ernst and Kellis 2010, 2012), take functional genomics data as input (DNase-seq; FAIRE-seq; and ChIP-seq of histonemodifications, RNA polymerase II large subunit [POLR2A], and CTCF) and return segmentation classes, wh”
Ensembl 2015 (2014, methodology): “To coincide with the annotation of the GRCh38 assembly, we inaugurated a new annotation algorithm (Ensembl Regulation Resources, in preparation) that computes regions of interest from genomic segmentations (20,21).”