labjc / pydeg

Differential expression analysis for bulk RNA-seq data.

PyDEG is a python implementation of the differential expression analysis(DEA) with bulk RNA-seq data. You have to note that it is just a mimic of DESeq2 of python and DESeq2 of R. We are neither majors nor experts.

Please refer to the following links : https://github.com/owkin/PyDESeq2 & https://hbctraining.github.io/DGE_workshop/lessons/04_DGE_DESeq2_analysis.html

Currently, PyDEG uses median of ratios as the size factors, like DESeq2. However, there are many differeces between PyDEG and DESeq2(or PyDESeq2). For instance, We use the likelihood-ratio test unlike DESeq2 which mainly uses the wald test.

There may be many errors and the code may reflect things we know are wrong. We will fix these things. If you notice anything strange, please feel free to contact me at any time.

Contents

1. Installation
2. How to use
3. License

---

Installation

PyDEG can be installed from PyPI using pip

pip install PyDEG

---

How to use

Import pydeg_dataset and results from pydeg. You may need import pandas.

import pandas as pd
from pydeg.pdd import pydeg_dataset, results

Create pydeg_dataset object where count_data is a pandas.DataFrame containing raw counts, metadata is a pandas.DataFrame containing information about samples and the design is string design formula.

pdd = pydeg_dataset(count_data, metadata, design)

Estimate size factors, dispersions and fit GLM for each gene.

pdd.DEA()

Test for differential expression with contrast = [condition, control, treat]

res = results(pdd, contrast)