DEPy

A differential expression analysis package for bulk proteomics (and metabolomics) data, which leverages transcriptomics tools. Inspired by R tools like DEP and SummarizedExperiment, it brings the power of Bioconductor to Python. All you need is a matrix of features and their intensity values.

• PyPI package: https://pypi.org/project/summarizedpy/
• GitHub: SebastianDHA/DEPy
• Free software: MIT License

Features

• SummarizedPY: A container for your -omics data, much like SummarizedExperiment or DEP in R.
• Filtering and subsetting your samples and features
• Missing value filtering
• Imputation using ImputeLCMD (many methods)
• Transforming (log, centering, standardizing, vsn)
• Leverage surrogate variable analysis (sva) to adjust for latent batch effects
• Use the flexibility and power of limma-trend to improve your DEA results and accommodate mixed effects
• Limma arrayWeights to adjust variable sample quality (often an issue in human and animal datasets)
• Visualize your DEA results with elegant volcano plots
• Highly-variable feature selection
• PCA plots
• Saving & loading SummarizedPy objects to & from disk

Installation

conda

This is the best way to install DEPy.

conda env create -f environment.yml

Note that DEPy (summarizedpy) must be run within the depy conda environment or a cloned version of it. This is because summarizedpy needs an isolated environment to run R in due to the complex loading behavior of Bioconductor packages.

Using pip

pip install summarizedpy

Quick start

import depy as dp

sp = dp.SummarizedPy()
sp = sp.import_from_delim_file(path="path/to/pgroup.tsv", delim="\t")

See the full tutorial for more.

Documentation

• GitHub pages
• ReadTheDocs

Credits

This package leverages amazing packages from the R and Bioconductor community, including limma, vsn, sva, ImputeLCMD, and Tidyverse. This package was created with Cookiecutter and the audreyfeldroy/cookiecutter-pypackage project template.