FEATURES

Automatic Annotation

A global reference

CellTypist recapitulates cell type structure and biology of independent datasets.

Robust models

Regularised linear models with Stochastic Gradient Descent provide fast and accurate prediction.

Easy to use

Scalable and flexible. Python-based implementation easy to integrate into existing pipelines.

Knowledge base

A community driven encyclopedia for cell types.

Get Started

Try it online

We'll send the results to your email.
Detail of each model can be found here .
Check how majority voting works here .
Only .csv and .h5ad file types accepted for online analysis. Desired format can be found here .

Install

With Python 3.6+ installed, get releases on PyPI or bioconda.

pip install celltypist

or

conda install -c bioconda -c conda-forge celltypist

Usage

Try the Tutorial Notebook alternative

Read the complete Usage Guide on the GitHub repo.

The typical workflow consists of selecting an input file and running the annotation process.

import celltypist
input_file = celltypist.samples.get_sample_csv()
#Run cell typing followed by majority voting.
predictions = celltypist.annotate(input_file, majority_voting = True)

#Examine the prediction results.
predictions.predicted_labels

Install

With Python 3.6+ installed, get releases on PyPI or bioconda.

pip install celltypist

or

conda install -c bioconda -c conda-forge celltypist

Usage

You can run CellTypist from anywhere on the command line.

celltypist --indata /path/to/input_data --majority-voting