Picking ChIP-seq peak detectors for analyzing chromatin modification experiments

Micsinai, Mariann; Parisi, Fulvio; Strino, Francesco; Asp, Patrik; Dynlacht, Brian David; Kluger, Yuval

doi:10.1093/nar/gks048

Cited by 66 publications

(74 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To date, a validated, complete epigenomic data set does not exist against which a software tool can be tested. However, several unbiased validation sets are available and it is advisable to choose a tool ranked high in a performance comparison with such data sets (Micsinai et al, 2012).…”

Section: Epigenetics and Transcriptional (Dys) Regulation In Diseasedmentioning

confidence: 99%

Epigenetics in the Human Brain

Houston

Peter

Mitchell

et al. 2012

Neuropsychopharmacol

View full text Add to dashboard Cite

Many cellular constituents in the human brain permanently exit from the cell cycle during pre-or early postnatal development, but little is known about epigenetic regulation of neuronal and glial epigenomes during maturation and aging, including changes in mood and psychosis spectrum disorders and other cognitive or emotional disease. Here, we summarize the current knowledge base as it pertains to genome organization in the human brain, including the regulation of DNA cytosine methylation and hydroxymethylation, and a subset of (altogether 4100) residue-specific histone modifications associated with gene expression, and silencing and various other functional chromatin states. We propose that high-resolution mapping of epigenetic markings in postmortem brain tissue or neural cultures derived from induced pluripotent cells (iPS), in conjunction with transcriptome profiling and whole-genome sequencing, will increasingly be used to define the molecular pathology of specific cases diagnosed with depression, schizophrenia, autism, or other major psychiatric disease. We predict that these highly integrative explorations of genome organization and function will provide an important alternative to conventional approaches in human brain studies, which mainly are aimed at uncovering group effects by diagnosis but generally face limitations because of cohort size.

show abstract

Section: Epigenetics and Transcriptional (Dys) Regulation In Diseasedmentioning

confidence: 99%

Epigenetics in the Human Brain

Houston

Peter

Mitchell

et al. 2012

Neuropsychopharmacol

View full text Add to dashboard Cite

show abstract

“…Central banks combine surveys of several professional forecasters to monitor rates of inflation, real gross domestic product growth, and unemployment (3)(4)(5)(6). Biologists study the genomic binding locations of proteins by combining or ranking the predictions of several peak detection algorithms applied to large-scale genomics data (7). Physician tumor boards convene a number of experts from different disciplines to discuss patients whose diseases pose diagnostic and therapeutic challenges (8).…”

mentioning

confidence: 99%

Ranking and combining multiple predictors without labeled data

Parisi

Strino

Nadler

et al. 2014

Proc. Natl. Acad. Sci. U.S.A.

Self Cite

106

View full text Add to dashboard Cite

In a broad range of classification and decision-making problems, one is given the advice or predictions of several classifiers, of unknown reliability, over multiple questions or queries. This scenario is different from the standard supervised setting, where each classifier's accuracy can be assessed using available labeled data, and raises two questions: Given only the predictions of several classifiers over a large set of unlabeled test data, is it possible to (i) reliably rank them and (ii) construct a metaclassifier more accurate than most classifiers in the ensemble? Here we present a spectral approach to address these questions. First, assuming conditional independence between classifiers, we show that the offdiagonal entries of their covariance matrix correspond to a rank-one matrix. Moreover, the classifiers can be ranked using the leading eigenvector of this covariance matrix, because its entries are proportional to their balanced accuracies. Second, via a linear approximation to the maximum likelihood estimator, we derive the Spectral Meta-Learner (SML), an unsupervised ensemble classifier whose weights are equal to these eigenvector entries. On both simulated and real data, SML typically achieves a higher accuracy than most classifiers in the ensemble and can provide a better starting point than majority voting for estimating the maximum likelihood solution. Furthermore, SML is robust to the presence of small malicious groups of classifiers designed to veer the ensemble prediction away from the (unknown) ground truth.spectral analysis | classifier balanced accuracy | unsupervised learning | cartels | crowdsourcing E very day, multiple decisions are made based on input and suggestions from several sources, either algorithms or advisers, of unknown reliability. Investment companies handle their portfolios by combining reports from several analysts, each providing recommendations on buying, selling, or holding multiple stocks (1, 2). Central banks combine surveys of several professional forecasters to monitor rates of inflation, real gross domestic product growth, and unemployment (3-6). Biologists study the genomic binding locations of proteins by combining or ranking the predictions of several peak detection algorithms applied to large-scale genomics data (7). Physician tumor boards convene a number of experts from different disciplines to discuss patients whose diseases pose diagnostic and therapeutic challenges (8). Peer-review panels discuss multiple grant applications and make recommendations to fund or reject them (9). The examples above describe scenarios in which several human advisers or algorithms provide their predictions or answers to a list of queries or questions. A key challenge is to improve decision making by combining these multiple predictions of unknown reliability. Automating this process of combining multiple predictors is an active field of research in decision science (cci.mit.edu/research), medicine (10), business (refs. 11 and 12 and www.kaggle.com/competitions), and government...

show abstract

“…Recent reports suggest that there remains no single gold-standard methodology for defining intervals of interest in epigenomic data sets and this could impact the distinction of chromatin marks. 17,25,26 Contrasting peak-calling and sliding window analyses, our de novo method uses assembled contigs, thereby avoiding the need for arbitrary selection or optimization of interval size and shape parameters.…”

Section: Discussionmentioning

confidence: 99%

Non-referenced genome assembly from epigenomic short-read data

Kaspi¹,

Ziemann

Keating

et al. 2014

Epigenetics

View full text Add to dashboard Cite

Abbreviations: ChIP-seq, immunoprecipitated chromatin sequencing; RNA-seq, RNA sequencing; MeDIP-seq, methylated DNA immunoprecipitation sequencing; MBD-seq, methyl binding domain protein sequencing; DMR, differentially methylated region.Current computational methods used to analyze changes in DNA methylation and chromatin modification rely on sequenced genomes. Here we describe a pipeline for the detection of these changes from short-read sequence data that does not require a reference genome. Open source software packages were used for sequence assembly, alignment, and measurement of differential enrichment. The method was evaluated by comparing results with reference-based results showing a strong correlation between chromatin modification and gene expression. We then used our de novo sequence assembly to build the DNA methylation profile for the non-referenced Psammomys obesus genome. The pipeline described uses open source software for fast annotation and visualization of unreferenced genomic regions from short-read data.

show abstract

Picking ChIP-seq peak detectors for analyzing chromatin modification experiments

Cited by 66 publications

References 32 publications

Epigenetics in the Human Brain

Epigenetics in the Human Brain

Ranking and combining multiple predictors without labeled data

Non-referenced genome assembly from epigenomic short-read data

Contact Info

Product

Resources

About