DECNEO

This repository contains DECNEO, a Python package that provides bioinformatics utilities for analyzing single cell transcriptomics datasets. DECNEO implements in silico detection of transcriptional regulation genes. The documentation is available at Read the Docs: https://decneo.readthedocs.io/

• Getting Started
  • Installation
  • Dependencies
• Functionality
  • Overview
  • Input Data Format
  • Usage Example
  • Output
• Funding
• Licensing

Getting Started

These are the instructions on how to get a copy of this project up and use it for data analysis.

Installation

The software runs in Python >= 3.8

To install DECNEO as a package:

$ pip install decneo

Alternatively, clone a local copy of this project to install the package from the cloned directory:

git clone https://github.com/sdomanskyi/decneo
python setup.py install

Dependencies

DECNEO is dependent on the following packages, that are installed/updated with installation of DECNEO:

• Matplotlib - plotting from Python
• NetworkX - used in network enrichment analysis
• Pandas and tables - for data storage and analysis
• NumPy - for processing data
• sklearn - we use clustering algorithms and metrics
• adjustText - optimization of text labels locations in plots

Functionality

Overview

The main implementation of DECNEO includes workflow for fast and efficient calculation of single cell gene expression distance (e.g. correlation) followed by the bootstrap technique to account for variation and noise in the input data. The results are summarized in a form of a optimized dendrogram, heatmap and information panels. Analysis of combination of measurements panels allows to identify main and secondary groups of genes that are coexpressed in the cell type of interest.

Input Data Format

Expression data for two different species for comparison is required. For each of these species provide the input gene expression data is expected in one of the following formats:

1. Spreadsheet of comma-separated values csv where rows are genes, columns are cells with gene expression counts, this should be accompanied by another dataframe with two columns with one specifying batches and the other specifying corresponding cells. Alternatively, the first row of the dataframe should be 'batch' and the second 'cell'.

2. Pandas DataFrame where axis 0 is genes and axis 1 are cells. If the are batched in the data then the index of axis 1 should have two levels, e.g. ('batch', 'cell'), with the first level indicating patient, batch or expreriment where that cell was sequenced, and the second level containing cell barcodes for identification.

For examples refer to documentation.

Usage Example

We have made an example execution file demo.py that shows how to use decneo.

Download file VoightChoroid4567RemappedData.h5 (456.7 Mb) from https://doi.org/10.5281/zenodo.4419880

This file contains normalized gene expression of 27504 genes of 7996 endothelial cells from 8 batches, and 5704 non-endothelial cells from 8 batches. Genes that are not expressed in endothelial cells are removed from non-endothelial cells dataset

Save the downloaded data file to demo/, or otherwise modify path in demoData of demo.py:

See details of the script demo.py at:

Example walkthrough of demo.py script

To execute the complete script demo.py run:

python demo.py

If reading demo data gives error "unsupported pickle protocol: 5" make sure that python 3.8 is used and latest version of pandas and tables is installed.

Output

Outputs all resulting directories, files, and figures to directory specified as the workingDir when creating an instance of class Analysis. It will also output an analysis report detailing all results and figures.

For a detailed list, refer to the documentation.

Funding

This research project is a part of R01GM122085 grant, funded by NIH/NIGMS.

Licensing

DECNEO is released under an MIT License. Please also consult the folder LICENSES distributed with DECNEO regarding Licensing information for use of external associated content.