Tutorials

The easiest way to get familiar with pertpy is to follow along with our tutorials. Many are also designed to work seamlessly in Google colab.

Note

For questions about the usage of pertpy use the scverse discourse.

Quick start: Tool specific tutorials

Data transformation

Guide RNA assignment
Mixscape - analysis of single-cell pooled CRSIPR screen
Perturbation Space
Metadata annotation
Ontology mapping

Knowledge inference

Augur - cell type prioritization prediction
scCODA - Compositional analysis of labeled single-cell data
scCODA - Modeling options and result analysis
tascCODA - Tree-aggregated compositional analysis
Milo - KNN based differential abundance analysis
DIALOGUE - multi cellular programs
Enrichment
Distance metrics
Distance Tests
CINEMA-OT
scGen - Perturbation response prediction
Differential gene expression

Use cases

Our use cases showcase a variety of pertpy tools applied to one dataset. They are designed to give you a sense of how to use pertpy in a real-world scenario. The use cases featured here are those we present in the pertpy preprint.

Use-case: Exploring genetic interation manifolds
Use-case: Deconvoluting drug responses in cancer cell lines
Use-case: Exploring perturbation effects

Glossary

AnnData is short for Annotated Data and is the primary datastructure that pertpy uses. It is based on the principle of a single Numpy matrix X embraced by two Pandas Dataframes. All rows are called observations (in our case cells or similar) and the columns are known as variables (any feature such as e.g. genes or similar). For a more in depth introduction please read the AnnData paper.

For a more in depth introduction please read the Scanpy paper.