Related Experiment Videos
Explicit DNase sequence bias modeling enables high-resolution transcription factor footprint detection
Galip Gürkan Yardımcı1, Christopher L Frank2, Gregory E Crawford3
1Computational Biology and Bioinformatics Program, Duke University, Durham, NC 27708, USA Center for Genomic and Computational Biology, Duke University, Durham, NC 27708, USA.
Nucleic Acids Research
|October 9, 2014
Summary
This study corrects for DNase cleavage bias in DNase-seq data to accurately identify transcription factor (TF)-DNA interactions. The new models improve prediction of TF binding sites and reveal insights into TF binding dynamics.
Area of Science:
- Genomics
- Molecular Biology
- Bioinformatics
Background:
- DNaseI footprinting identifies transcription factor (TF)-DNA interactions.
- High-throughput DNase-seq detects in vivo DNase footprints genome-wide.
- Existing computational methods for DNase-seq footprinting are impacted by DNase cleavage bias.
Purpose of the Study:
- To assess DNase-seq for identifying individual TF binding sites.
- To develop bias-corrected, TF-specific footprint models.
- To evaluate the predictive performance of these models.
Main Methods:
- Performed DNase-seq on deproteinized genomic DNA to determine sequence cleavage bias.
- Developed bias-corrected and TF-specific footprint models.
- Validated predicted footprints against high-confidence TF-DNA interactions and ChIP-seq data.
Main Results:
- Bias-corrected models accurately predict TF-DNA interactions.
- Absence of footprints under some ChIP-seq peaks indicates weaker or indirect binding, or ChIP artifacts.
- Detected variations in TF binding motifs and cell-type-specific footprints within DNase hypersensitive sites.
Conclusions:
- DNase-seq, when corrected for cleavage bias, is a powerful tool for identifying TF binding sites.
- This approach reveals nuances in TF binding not detectable by other methods.
- The models provide a more accurate understanding of TF-DNA interactions and binding dynamics.
