Combining transcription factor binding affinities with open-chromatin data for accurate gene expression prediction

Schmidt, Florian; Gasparoni, Nina; Gasparoni, Gilles; Gianmoena, Kathrin; Cadenas, Cristina; Polansky, Julia K.; Ebert, Peter; Nordström, Karl; Barann, Matthias; Sinha, Anupam; Fröhler, Sebastian; Xiong, Jieyi; Amirabad, Azim Dehghani; Ardakani, Fatemeh Behjati; Hutter, Barbara; Zipprich, Gideon; Felder, Bärbel; Eils, Jürgen; Brors, Benedikt; Chen, Wei; Hengstler, Jan G.; Hamann, Alf; Lengauer, Thomas; Rosenstiel, Philip; Walter, Jörn; Schulz, Marcel H.

doi:10.1093/nar/gkw1061

Cited by 111 publications

(107 citation statements)

References 72 publications

Supporting

Mentioning

104

Contrasting

Order By: Relevance

“…A more refined estimation of TF activity can be obtained with combined analysis of open chromatin (e.g. ATAC-seq and DNAseI-Seq) and TFBS over-representation [38]. Over-representation analysis based on publicly available annotated genes is an unsupervised, fast and usually helpful first step of gene cluster interpretation.…”

Section: (C) Limitations and Challengesmentioning

confidence: 99%

Assessment of stem cell differentiation based on genome-wide expression profiles

Godoy

Schmidt‐Heck

Hellwig

et al. 2018

Phil. Trans. R. Soc. B

Self Cite

View full text Add to dashboard Cite

In recent years, protocols have been established to differentiate stem and precursor cells into more mature cell types. However, progress in this field has been hampered by difficulties to assess the differentiation status of stem cell-derived cells in an unbiased manner. Here, we present an analysis pipeline based on published data and methods to quantify the degree of differentiation and to identify transcriptional control factors explaining differences from the intended target cells or tissues. The pipeline requires RNA-Seq or gene array data of the stem cell starting population, derived 'mature' cells and primary target cells or tissue. It consists of a principal component analysis to represent global expression changes and to identify possible problems of the dataset that require special attention, such as: batch effects; clustering techniques to identify gene groups with similar features; over-representation analysis to characterize biological motifs and transcriptional control factors of the identified gene clusters; and metagenes as well as gene regulatory networks for quantitative cell-type assessment and identification of influential transcription factors. Possibilities and limitations of the analysis pipeline are illustrated using the example of human embryonic stem cell and human induced pluripotent cells to generate 'hepatocyte-like cells'. The pipeline quantifies the degree of incomplete differentiation as well as remaining stemness and identifies unwanted features, such as colon- and fibroblast-associated gene clusters that are absent in real hepatocytes but typically induced by currently available differentiation protocols. Finally, transcription factors responsible for incomplete and unwanted differentiation are identified. The proposed method is widely applicable and allows an unbiased and quantitative assessment of stem cell-derived cells.This article is part of the theme issue 'Designer human tissue: coming to a lab near you'.

show abstract

Section: (C) Limitations and Challengesmentioning

confidence: 99%

Assessment of stem cell differentiation based on genome-wide expression profiles

Godoy

Schmidt‐Heck

Hellwig

et al. 2018

Phil. Trans. R. Soc. B

Self Cite

View full text Add to dashboard Cite

show abstract

“…The DNase-seq assay has been developed to measure chromatin accessibility genome-wide (Sabo et al 2004;Crawford et al 2006;John et al 2013). "Footprinting" algorithms attempt to recover cell-specific TF binding sites (TFBSs) from DNase cleavage signals (Zhang et al 2008;Hesselberth et al 2009;Boyle et al 2011;Pique-Regi et al 2011;Cuellar-Partida et al 2012;Piper et al 2013;Gusmao et al 2014Gusmao et al , 2016Sherwood et al 2014;Sung et al 2014;Yardımcı et al 2014;Kähärä and Lähdesmäki 2015;Chen et al 2017;Schmidt et al 2017;Quach and Furey 2017).…”

mentioning

confidence: 99%

Anchor: trans-cell type prediction of transcription factor binding sites

Quang

2018

Genome Res.

View full text Add to dashboard Cite

The ENCyclopedia of DNA Elements (ENCODE) consortium has generated transcription factor (TF) binding ChIP-seq data covering hundreds of TF proteins and cell types; however, due to limits on time and resources, only a small fraction of all possible TF-cell type pairs have been profiled. One solution is to build machine learning models trained on currently available epigenomic data sets that can be applied to the remaining missing pairs. A major challenge is that TF binding sites are cell-type–specific, which can be attributed to cellular contexts such as chromatin accessibility. Meanwhile, indirect TF-DNA binding and interactions between TFs complicate this regulatory process. Technical issues such as sequencing biases and batch effects render the prediction task even more challenging. Many pioneering efforts have been made to predict TF binding profiles based on DNA sequence and DNase-seq footprints, but to what extent a model can be generalized to completely untested cell conditions remains unknown. In this study, we describe our first place solution to the 2017 ENCODE-DREAM in vivo TF binding site prediction challenge. By carefully addressing multisource biases and information imbalance across cell types, we created a pipeline that significantly outperforms the current state-of-the-art methods. The proposed method is sufficiently complex enough to model nonlinear interactions between TF binding motifs and chromatin accessibility information up to 1500 bp from the genomic region of interest.

show abstract

“…Importantly, however, information about cell state transitions, the history of a cell, and how temporal events regulate cellular transitions is lost or hidden. Nonetheless, powerful computational inference frameworks have emerged that support the move from descriptive studies of cellular states to models of dynamic events (Bendall et al, 2014;Chen et al, 2016b;Guo et al, 2017;Haghverdi et al, 2016;Herring et al, 2018;Qiu et al, 2017;Setty et al, 2019;Setty et al, 2016;Shin et al, 2015;Trapnell et al, 2014;Weinreb et al, 2018;Wolf et al, 2019). These methods assume that single cells transit from one cellular state to another in a continuous fashion, and that all necessary cellular states for the process under investigation are sampled with sufficient depth, allowing the ordering of cells along a pseudotime trajectory of cellular progression.…”

Section: Temporally Resolved Single Cell Methodsmentioning

confidence: 99%

“…These methods assume that single cells transit from one cellular state to another in a continuous fashion, and that all necessary cellular states for the process under investigation are sampled with sufficient depth, allowing the ordering of cells along a pseudotime trajectory of cellular progression. This process of 'trajectory inference' has been applied successfully to various imaging (Gut et al, 2015;Herring et al, 2018;Serra et al, 2019), CyTof (Bendall et al, 2014;Setty et al, 2016) and sequencing (Chen et al, 2016b;Guo et al, 2017;Haghverdi et al, 2016;Qiu et al, 2017;Setty et al, 2019;Shin et al, 2015;Trapnell et al, 2014;Weinreb et al, 2018;Wolf et al, 2019) datasets. However, trajectory inference solely on cellular states has its limitations, as reviewed recently elsewhere (Kester and van Oudenaarden, 2018;Wagner et al, 2016;and, also in this issue, Tritschler et al, 2019).…”

Section: Temporally Resolved Single Cell Methodsmentioning

confidence: 99%

Exploring single cells in space and time during tissue development, homeostasis and regeneration

2019

View full text Add to dashboard Cite

Complex 3D tissues arise during development following tightly organized events in space and time. In particular, gene regulatory networks and local interactions between single cells lead to emergent properties at the tissue and organism levels. To understand the design principles of tissue organization, we need to characterize individual cells at given times, but we also need to consider the collective behavior of multiple cells across different spatial and temporal scales. In recent years, powerful single cell methods have been developed to characterize cells in tissues and to address the challenging questions of how different tissues are formed throughout development, maintained in homeostasis, and repaired after injury and disease. These approaches have led to a massive increase in data pertaining to both mRNA and protein abundances in single cells. As we review here, these new technologies, in combination with in toto live imaging, now allow us to bridge spatial and temporal information quantitatively at the single cell level and generate a mechanistic understanding of tissue development.

show abstract

Combining transcription factor binding affinities with open-chromatin data for accurate gene expression prediction

Cited by 111 publications

References 72 publications

Assessment of stem cell differentiation based on genome-wide expression profiles

Assessment of stem cell differentiation based on genome-wide expression profiles

Anchor: trans-cell type prediction of transcription factor binding sites

Exploring single cells in space and time during tissue development, homeostasis and regeneration

Contact Info

Product

Resources

About