fadhlyemen/doubletdetection

Doublet detection in single-cell RNA-seq data.

DoubletDetection

DoubletDetection is a Python3 package to detect doublets (technical errors) in Single-cell RNA-seq count tables.

To install DoubletDetection, clone or download this repository.

To run DoubletDetection you may need to install the following packages:

• numpy
• pandas
• sklearn
• phenograph

To install PhenoGraph:

pip3 install git+https://github.com/jacoblevine/phenograph.git

To run basic doublet classification:

import doubletdetection
raw_counts = doubletdetection.load_csv('/your/path/filename.csv')
clf = doubletdetection.BoostClassifier()
labels = clf.fit(raw_counts)

raw_counts is a scRNA-seq count table. labels is a binary vector with the value 1 representing a detected doublet. The length of labels is equal to the length of raw_counts.

Advanced usage:

See our jupyter notebook.

Obtaining data

Data can be downloaded from the 10x website.

This project is licensed under the terms of the MIT license.