Panpipes and the scverse

Panpipes is an analytical pipeline implemented in python, which orchestrates single cell analyses in an automated fashion. Panpipes relies on the scverse, an ecosystem of tools for single cell ‘omics analysis in python. Checkout the scverse documentation for the packages included and how to contribute your package so we can support it in panpipes too!

Panpipes has at its core AnnData and MuData, for handling annotated data matrices in memory and on disk, with the best of the pandas and xarray functionalities.

`AnnData` is anndata is a container for handling annotated data matrices objects.

`MuData` is a dictionary of `AnnData` objects.

The workhorses for panpipes are scanpy, muon and squidpy, frameworks for analyzing single-cell gene expression, multimodal data and spatial trascriptomics.

For deep-learning based methods for uni or multimodal integration, we leverage the functionalities of scvi-tools, a library developing probabilistic models for single-cell omics data in PyTorch.

To get help with panpipes, you can open an issue on our Github page. Please use the scverse discourse to document issues with scverse packages and get the help of other scverse users!

Please use these links to familiarize with these data structures and frameworks:

• AnnData

• MuData Quickstart

• Single cell analysis with scanpy

• Multimodal analyses with muon

• Spatial analyses with squidpy

• Scvi-tools

• Interoperability

• Best-Practices for single cell multimodal analyses