Stardust

Python package for cell-gene coembedding and analysis pipeline for single cell transcriptomics data

Installation
Dataset Format
Usage
Tutorial
- Co-embedding
Analysis

Environment reqirements

Pyhton 3.7 or above with setuptools instaled.

Installation Guide

Download the package directly from github. For linux users

https://github.com/Swagatam123/Stardust_package.git

Download the binary openOrd.jar from the release tag v1.0 and place it under Stardust_package/stardust/run_stardust/ Change the directoery to the package directory

cd Stardust_package/

Use the package manager pip to install package.

pip install .

Dataset support fomrat

The package supports 10x genomics data with .10x file format and csv file with the filename expression.csv The expression data should contain the cells rowwise and genes columnwise.

Usage

import stardust

stardust.run_stardust.run() #generated the coembedding plots under Stardust_resuts/visualization_output/4_pass.
stardust.analysis.silhouette() # optional for analysis purpose
stardust.analysis.heatmap()   # optional for analysis purpose
stardust.analysis.alluvial()  # optional for analysis purpose

Vignette tutorial

This vignette uses a melanoma data set from the website here to demonstrate a standard pipeline. This vignette can be used as a tutorial as well.

cell-gene co-embedding

To re-produce the cell-gene co-embedding visualization you need to run the following commands

import stardust

stardust.run_stardust.run() #generated the coembedding plots under Stardust_resuts/visualization_output/4_pass.

On run command, you need to provide the expression matrix path and the file format (i.e either "csv" or "10x" ). NOTE: For csv file format, the expression file name should be "expression.csv"