Introduction

cPeaks serves as a unified reference for different cell types for scATAC-seq data, enhancing downstream analysis, particularly in cell annotation and rare cell type detection.

cPeaks Download

We released cPeaks (.bed format) with GRCh37/hg19 and GRCh38/hg38 version.

The basic information and properties of cPeaks can be found in cPeaks_info.tsv.

cPeaks Usage

Quick Start Guide

cPeaks eliminates the need for the peak-calling step by providing a ready-to-use reference. If you are familiar with SnapATAC2, ArchR or Python, this section will help you quickly get started with cPeaks. However, if you have any questions, please refer to detailed tutorials.

• SnapATAC2

  # Read cPeaks file cpeaks_hg19.bed or cpeaks_hg38.bed
  cpeaks_path = 'YOUR_PATH/cpeaks_hg38.bed'
  with open(cpeaks_path) as cpeaks_file:
    cpeaks = cpeaks_file.read().strip().split('\n')
    cpeaks = [peak.split('\t')[0] + ':' + peak.split('\t')[1] + '-' + peak.split('\t')[2] for peak in cpeaks]
  # Set parameter use_rep to 'cpeaks' as mapping reference. 'data' is your AnnData object.
  data = snap.pp.make_peak_matrix(data, use_rep=cpeaks)

• ArchR

  # Read cPeaks file cpeaks_hg19.bed or cpeaks_hg38.bed
  cpeaks <- read_table('YOUR_PATH/cpeaks_hg19.bed', col_names = F)
  cpeaks.gr <- GRanges(seqnames = cpeaks$X1, ranges = IRanges(cpeaks$X2, cpeaks$X3))
  # Set parameter 'features' to cpeaks.gr. 'proj' is your ArchRProject object.
  proj <- addFeatureMatrix(proj, features = cpeaks.gr, matrixName = 'FeatureMatrix')

• Run Python Script Manually

  git clone https://github.com/MengQiuchen/cPeaks.git
  cd cPeaks/map2cpeak
  python main.py --fragment_path PATH/to/YOUR_fragment.tsv.gz --output map2cpeaks_result

  After running main.py as above, you will get a matrix file cell_cpeaks.mtx and a text file barcodes.txt within a newly created folder named map2cpeaks_result.

  Learn more about the arguments for main.py.

Comprehensive Guide

Please refer to detailed tutorials.

Citation

Meng Q, Wu X, et al. Toward a generic feature set defined by consensus peaks as a consistent reference for ATAC-seq data. Preprint at bioRxiv https://doi.org/10.1101/2023.05.30.542889 (2023).

Contact

Please reach out to Meng Qiuchen at mqc17@mails.tsinghua.edu.cn if you encounter any issues or have any recommendations.