/BandNorm_and_scVI-3D_manuscript

Code for generating figures and tables from "Normalization and De-noising of Single-cell Hi-C Data with BandNorm and scVI-3D"

Primary LanguageJupyter NotebookGNU General Public License v3.0GPL-3.0

Codes for Manuscript

Reproducible codes used for conducting analysis and generating figures and tables of the manuscript Normalization and de-noising of single-cell Hi-C data with BandNorm and scVI-3D. BioRxiv (2021). Ye Zheng*, Siqi Shen* and Sündüz Keleş. * contribute equally.

The software source codes are deposited at

The curated scHi-C data used in this manuscript are cleaned and organized at https://pages.stat.wisc.edu/~sshen82/bandnorm/ for easy access and direct usage in BandNorm and scVI-3D. This online data repository contains five scHi-C data sets, all at 1Mb. The dataset name, raw data source and reference are summarized below:

The metadata summaries that contains the cell barcode, cell type, sequencing depth and sparsity for each data set are also provided at https://pages.stat.wisc.edu/~sshen82/bandnorm/Summary/.

In this repository, the scripts are classified into three subfolders based on the purpose.

  • RunAlgorithms folder contains scripts to run BandNorm, scVI-3D, scHi-C normalization baseline methods that we created for comparison and other publically available tools including scHiCluster, scHiC Topics and Higashi. The outputs of those scripts provide lower-dimensional embeddings and normalized scHi-C counts, which will be used for the downstream analysis and generating figures.

  • AnalysisCodes folder contains analysis related scripts to implement evaluation, comparison and biological discoveries.

  • FigureGeneration folder contains scripts that generate the figures in the manuscript.

For questions, please email Siqi Shen (sshen82@wisc.edu) or submit new issue at this repository.

This BandNorm_and_scVI-3D_manuscript repository is protected by GNU GPL 3 license. Please refer to LICENSE.md for using codes in this repository for your own purposes.