---
title: cpa
description: The Compositional Perturbation Autoencoder (CPA) is a deep generative
  framework to learn effects of perturbations at the single-cell level. CPA performs
  OOD predictions of unseen combinations of drugs, learns interpretable embeddings,
  estimates dose-response curves, and provides uncertainty estimates.
tags:
- human-cell-atlas
- perturbation
- scrna-seq
- single-cell
- single-cell-genomics
- single-cell-rna-seq
- source-code
license: BSD-3-Clause
source_url: https://github.com/theislab/cpa
---

# cpa

> **Mirrored metadata.** This card is reproduced on Atlas for discovery. The artifact itself is not hosted here — follow the source link for the files, and refer to the upstream licence for terms of use.

The Compositional Perturbation Autoencoder (CPA) is a deep generative framework to learn effects of perturbations at the single-cell level. CPA performs OOD predictions of unseen combinations of drugs, learns interpretable embeddings, estimates dose-response curves, and provides uncertainty estimates.

Upstream repository: `theislab/cpa`.

**Source:** https://github.com/theislab/cpa

**Licence:** BSD-3-Clause

_This model is part of the Atlas demo catalogue. See the repository history for provenance._