The identification of footprints provides a powerful means to detect base-pair resolution signals of regulatory elements in chromatin accessibility data. We observe substantial variability in Tn5 transposase cleavage bias across individual samples, highlighting the need for sample-specific correction. TraceBIND is a R package that that corrects sample-specific Tn5 bias through mitochondria-based fine-tuning of PRINT’s deep learning model, and use it to identify TF and nucleosome footprints through a dynamic flanking window statistical scan. TraceBIND also enables sample-specific FDR-controlled p-value thresholds stratified by varying coverage, which is necessary because coverage can vary across orders of magnitude in ATAC-seq data.
Before installing this package, make sure you have the following dependencies installed:
# install.packages("devtools")
devtools::install_github("yaowuliu/ACAT")TraceBIND has been tested on python=3.11, 3.12. The requirements of python packages for traceBIND finetuning are listed in the requirements, which can be done by:
pip install -r requirements.txtThen install this package:
# install.packages("devtools")
devtools::install_github("lyx-lin/TraceBIND", dependencies=TRUE)Please refer to the following jupyter notebooks for tutorials:
- Finetuning sample-specific Tn5 bias model on mito DNA (Before finetuning, please download PRINT Tn5 bias model)
- Identification of footprints
- Footprint-informed chromVar analysis
The data used could be found here.
For more model details, validation results and real dataset analysis, please check out our manuscript on bioRXiv. If you use our method, please use the following citation:
@article {Lin2025.10.17.683160,
author = {Lin, Yuxuan and Wang, Hanzhi and Wilson, Parker C. and Zhang, Nancy R.},
title = {Robust footprinting with sample-specific Tn5 bias correction for bulk and single cell ATAC-seq},
elocation-id = {2025.10.17.683160},
year = {2025},
doi = {10.1101/2025.10.17.683160},
publisher = {Cold Spring Harbor Laboratory},
URL = {https://www.biorxiv.org/content/early/2025/10/18/2025.10.17.683160},
journal = {bioRxiv}
}