Welcome to SpaCr

Docs PyPI version Python version Licence: MIT repo size Tutorial

spaCR (Spatial Phenotype Analysis of CRISPR Screens) is a Python toolkit for analyzing pooled CRISPR-Cas9 imaging screens. It integrates high-content imaging data with sequencing-based mutant identification to enable genotype-to-phenotype mapping at the single-cell level.

spaCR supports:

  • Segmentation of microscopy images using models like Cellpose.

  • Single-cell feature extraction and image cropping.

  • Classification of phenotypes using classical and deep learning models.

  • Barcode decoding from sequencing reads and well-level mutant quantification.

  • Statistical analysis, including regression models to link genotypes to phenotypes.

  • Interactive visualization of results including Grad-CAMs and phenotype maps.

  • GUI tools for mask curation, annotation, and exploratory analysis.

Example Notebooks

The following example Jupyter notebooks illustrate common workflows using spaCR.

API Reference by Category

Classification

Sequencing & Submodules

GitHub Repository

Visit the source code on GitHub: https://github.com/EinarOlafsson/spacr