Welcome to spaCR¶

spaCR — Spatial phenotype analysis of CRISPR screens.

Note

You are reading the stable documentation for spaCR 1.5.1.0. The main site follows main. The nightly preview follows nightly and may describe features not yet in a release. Each branch publishes its own API, guides and committed tutorial catalog automatically.

A Python toolkit for quantifying and visualising phenotypic changes in high-throughput microscopy screens. Ships with a modern PySide6 GUI (spacr), a headless pipeline (spacr.core), and a plate-to-classification workflow that runs on top of PyTorch, Cellpose, scikit-image, and scipy.

It is built for cell biologists running pooled or arrayed CRISPR screens who need per-cell measurements from plate images. The GUI route needs no programming; the same processing steps are available through the Python API for scripted and reproducible workflows.

The GUI groups its applications into four categories: Core for the segment-measure-classify pipeline, Data for getting images and tables in and running them at scale, Tools for the instruments you point at a project — hand mask correction, stitching, embeddings, gates, plots and quality control — and Assays for the Toxoplasma, Plasmodium and Candida organism guides. Choose an organism to see its available assays; planned analyses are marked Coming soon. The bands under “Applications and workflow” below are those categories, in that order, with the tiles each one holds.

Not every screen is a tile, which is why no count of them is printed here. Work that only makes sense inside another step opens from that step’s masthead instead — Timelapse from Mask, Illumination and the Motility Assay from Measure, Classifier Evaluation and Explain CV Model from Classify, Annotator Agreement from Annotate, and the Cellpose Workbench, Model Compare, Model Zoo and Curate from Make Masks, among them — so a screen with no tile below is one step further in rather than gone. Home lists whatever the running build offers.

🚀 Get started

Install spaCR from PyPI and launch the Qt GUI in two commands.

https://github.com/EinarOlafsson/spacr#install-spacr

🎓 Interactive tutorials

Narrated, step-by-step lessons for spaCR workflows and modules.

tutorials/

📖 API reference

Supported workflow entry points and the complete module reference.

API reference

🎬 Video tutorials

Narrated walkthroughs of each pipeline module.

tutorials/

🐛 Report an issue

File a bug, request a feature, or ask a question.

https://github.com/EinarOlafsson/spacr/issues/new

Applications and workflow¶

New to spaCR? Choose a workflow after installation for the first Home tile, the inputs each step needs, and what to open next.

Every tile links to the API page used by that application’s in-product help.

spaCR modules¶

Core¶

Core sequence from microscopy images through segmentation, measurements, annotations, classification, barcode mapping and regression.

Data¶

Import images and tables into spaCR projects and execute reproducible multi-plate workflows.

Tools¶

Point these at a project: edit masks by hand, stitch tiles, read an embedding, draw a gate, build a plot, check quality.

Assays¶

Quantitative readouts for biological assays.

Installation¶

Install spaCR and launch the desktop application. The qt extra remains supported for desktop installations; current packages also include PySide6 among their base dependencies.

python -m pip install "spacr[qt]"
spacr                    # launch the Qt GUI

For a terminal workflow on a cluster or server, use spacr-run without opening the desktop application:

python -m pip install spacr
spacr-run --list         # list the headless pipeline modules

Learn spaCR¶

Start with installation, continue to Home and the pipeline overviews, then follow the module walkthrough for your task in the interactive tutorial library. Each lesson lists its available narration voices and captions. New English lessons can appear while their translations are being prepared. Open the library from the GUI through Help → Tutorial (web).