Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 Aug 24;14(8):e1006372.
doi: 10.1371/journal.pcbi.1006372. eCollection 2018 Aug.

coTRaCTE predicts co-occurring transcription factors within cell-type specific enhancers

Alena van Bömmel  1 , Michael I Love  2 , Ho-Ryun Chung  3   4 , Martin Vingron  1
Affiliations

coTRaCTE predicts co-occurring transcription factors within cell-type specific enhancers

Alena van Bömmel et al. PLoS Comput Biol. .

Abstract

Cell-type specific gene expression is regulated by the combinatorial action of transcription factors (TFs). In this study, we predict transcription factor (TF) combinations that cooperatively bind in a cell-type specific manner. We first divide DNase hypersensitive sites into cell-type specifically open vs. ubiquitously open sites in 64 cell types to describe possible cell-type specific enhancers. Based on the pattern contrast between these two groups of sequences we develop "co-occurring TF predictor on Cell-Type specific Enhancers" (coTRaCTE) - a novel statistical method to determine regulatory TF co-occurrences. Contrasting the co-binding of TF pairs between cell-type specific and ubiquitously open chromatin guarantees the high cell-type specificity of the predictions. coTRaCTE predicts more than 2000 co-occurring TF pairs in 64 cell types. The large majority (70%) of these TF pairs is highly cell-type specific and overlaps in TF pair co-occurrence are highly consistent among related cell types. Furthermore, independently validated co-occurring and directly interacting TFs are significantly enriched in our predictions. Focusing on the regulatory network derived from the predicted co-occurring TF pairs in embryonic stem cells (ESCs) we find that it consists of three subnetworks with distinct functions: maintenance of pluripotency governed by OCT4, SOX2 and NANOG, regulation of early development governed by KLF4, STAT3, ZIC3 and ZNF148 and general functions governed by MYC, TCF3 and YY1. In summary, coTRaCTE predicts highly cell-type specific co-occurring TFs which reveal new insights into transcriptional regulatory mechanisms.

PubMed Disclaimer

Conflict of interest statement

The authors have declared that no competing interests exist.

Figures

Fig 1
Fig 1. Overview of the method for the determining of cell-type specific DNase hypersensitive sites (CTS-DHSs).
The top panel shows the raw DNase-seq data for 14 samples of 7 different cell lines (highlighted with different colors). Then, for each genomic window w1, …, w12, the log normalized read counts for each sample are calculated (matrix of read counts in the bottom right panel). The t-statistics are then calculated for each genomic window and each cell type over all corresponding cell lines (matrix of t-statistics in the bottom left panel). Windows w6 and w7 with large read counts in all cell types have small t-score over all cell types and are referred to as ubiquitously open DHSs. Windows w2 and w3 with large read counts in bone marrow have a large t-score in bone marrow only and are referred to as bone marrow-specific DHSs (e.g. CTS-DHSs).
Fig 2
Fig 2. Schematic of the method for the detection of TF overrepresentation and co-occurrence in the cell-type specific DNase hypersensitive sites (CTS-DHSs).
First, the l-most cell-type specific DHSs and the l-most ubiquitous DHSs are determined (1). Then, for each cell type separately, and for each TF of interest, CTS-DHSs and ubiquitous DHSs are jointly ranked by the binding affinity (2). The preset cutoff determines bound and unbound DHSs (3). This approach is repeated for all TF motifs and all investigated cell types (4). Co-occurring TF pairs on CTS-DHSs are predicted from the log score of p-values derived from two contingency tables with Fisher’s exact test (5) and summarized in cell-type specific TF networks (6) or in TF-specific networks with co-occurring partners in different cell types (7).
Fig 3
Fig 3. Cell-type specificity of predicted co-occurring TF pairs by CoTRaCTE.
A) Cumulative plot of number of cell types in which TF pairs are predicted to co-occur. As highlighted with the green lines, 70% of all predicted co-occurring TF pairs are highly cell-type specific i.e. they are observed in only 6 cell types or fewer. B) Heatmap of overlapping predicted co-occurring TF pairs over 64 cell types. Each cell depicts the number of TF pairs shared between the corresponding pair of cell types. Primary T-cell, HPCs and ESCs (highlighted with red dashed lines) possess sets of co-occurring TF pairs which are distinct from those of other cell types. Functionally related cell types such as microvascular endothelial dermal lymph cells and microvascular endothelial lung lymph cells (highlighted with green dashed lines) share a large number of TF pairs.
Fig 4
Fig 4. Comparison of predicted co-occurring TF pairs by CoTRaCTE with experimental data.
A) Proportion of experimentally validated PPIs in the predicted co-occurring TF pairs by CoTRaCTE by cell type, including the predicted TF pairs on ubiq-DHSs. Blue level indicates the proportion of PPIs in a random set of TF pairs. B) Comparison of interacting TF pairs predicted by ENCODE and by CoTRaCTE. Blue cords denote the high confidence set of interacting TFs from ENCODE, TF pairs predicted also by CoTRaCTE are highlighted in red. Experimental PPIs are highlighted with black border lines.
Fig 5
Fig 5. Network of predicted co-occurring TF pairs in undifferentiated A) and differentiated B) embryonic stem cells (H7-hESC cell line).
A) Network of predicted co-occurring TFs in ESCs. Nodes in the network represent transcription factors, edges are drawn between co-occurring TF pairs predicted by coTRaCTE. Red edges are known protein-protein interactions which are also predicted by coTRaCTE. TFs expressed in the cell line are highlighted in green; darker tone indicates stronger evidence of expression in related cell types. Known regulators in ESCs are highlighted as rectangles with yellow border. Node size reflects the number of predicted co-occurring TF partners. Three subnetworks with distinct regulatory functions are highlighted: maintanance of pluripotency (blue); embryonic development (pink) and general functions (orange). B) Network of predicted co-occurring TFs in differentiated ESCs. Four subnetworks with distinct regulatory functions are highlighted: pluripotency (blue); early development of organs and tissues (red); mesoderm differentiation (purple) and general functions (orange).
Fig 6
Fig 6. Predicted co-factors of GATA1 in different cell types.
Nodes in the network represent transcription factors, edges are drawn between the co-occurring TF pairs predicted by coTRaCTE, the width of each edge corresponds to number of cell types for which the TF co-occurrence was predicted. Coloured groups indicate different cell types for which the TF co-occurrence was predicted.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Maston Glenn A. and Evans Sara K. and Green Michael R.(2006). Transcriptional Regulatory Elements in the Human Genome. Annual Review of Genomics and Human Genetics, 7, 29–59. 10.1146/annurev.genom.7.080505.115623 - DOI - PubMed
    1. Zhu Zhou and Shendure Jay and Church George M. (2005). Discovering functional transcription-factor combinations in the human cell cycle. Genome Research, 15, 848–855. 10.1101/gr.3394405 - DOI - PMC - PubMed
    1. Hu Z. and Gallo S. (2010). Identification of interacting transcription factors regulating tissue gene expression in human. BMC Genomics, 11, 49+ 10.1186/1471-2164-11-49 - DOI - PMC - PubMed
    1. Myšičková A. and Vingron M. (2012). Detection of interacting transcription factors in human tissues using predicted DNA binding affinity. BMC Genomics, 13 Suppl 1, S2+ 10.1186/1471-2164-13-S1-S2 - DOI - PMC - PubMed
    1. Vandenbon A., Kumagai Y., Akira S., and Standley D. (2012). A novel unbiased measure for motif co-occurrence predicts combinatorial regulation of transcription. BMC Genomics, 13, S11+ 10.1186/1471-2164-13-S7-S11 - DOI - PMC - PubMed

Publication types

MeSH terms