mudi

A collection of python-based single-cell utilities and pipelines.

Author: Shankara Anand

Email: sanand@broadinstitute.org

For more information about mudis, the amazing dog breed this package is named after, please visit https://en.wikipedia.org/wiki/Mudi.

---

Requirements

• python>=3.6
• R>=3.5.1
  • Seurat>=3.0.0
  • scran>=1.10.1

Installation

git clone https://github.com/broadinstitute/mudi.git

cd mudi
pip install -e .

Support for PIP coming soon.

Package Details

Includes

• wrappers for sequence level QC (ie FastQC)
• pre-processing recipes for raw count data with qc & filtering helpers
• wrappers for single-cell normalization (i.e. scran)
• wrapper for signatureanalyzer (bayesian NMF)
• wrapper for gprofiler

Examples: see ./examples

• normalization: general processing & normalization
• cell_types: assigning cell-types
• ardnmf: running ard-nmf via signatureanalyzer

Example Usage

import mudi as md

adata = md.recipe(
  list_of_10x_files,
  min_genes=200,
  min_cells=3,
  compute_doublets=True,
  remove_doublets=False,
  verbose=True,
  norm='scran'
)

Examples and testing coming soon...

• Batch Integration
• Differential Expression
• Proportion Tests
• Plotting features
• QC plotting
• Integration with SCRINVEX