scPCA is a versatile matrix factorisation framework designed to analyze single-cell data across diverse experimental designs.
scPCA is a young project and breaking changes are to be expected.
scPCA enables the analysis of single-cell RNA-seq data across condtions. In simple words, it enables the incorporation of a design (model) matrix that encodes the experimental design of the dataset and infers how the gene loading weight vectors change from a specified reference condition to the treated condtion.
TwoVectorsPCAWithGreyCircles.mp4
scPCA makes use torch, pyro and anndata. We highly recommend to run scPCA on a GPU device.
The easiest option to install scpca is via Pypi. Simply type
$ pip install scpca
into your shell and hit enter.
- Free software: MIT license
- Documentation: https://sagar87.github.io/scPCA/index.html
- Harald Vöhringer