pyscalop 0.1.0

Single-cell analysis operations — Python port of scalop

pyscalop

Single-cell analysis operations — Python port of scalop.

A small toolbox for scRNA-seq analysis with a focus on intra-tumour expression programs (as in Gavish et al. 2023). Built on numpy, pandas, scipy, anndata.

Install

pip install -e .

Quick start

import pyscalop as ps

# m: pandas DataFrame, genes x cells, log-normalised, not row-centered
# sigs: dict of {signature_name: [gene1, gene2, ...]}
scores = ps.sig_scores(m, sigs)

# Differential expression for one or more cell groups vs the rest
deas = ps.dea(m, groups={"clusterA": cells_A, "clusterB": cells_B})

# Find intra-tumour programs (clustering + DEA + jaccard filtering)
res = ps.programs(m)

# Aggregate programs across samples into metaprograms
mp = ps.metaprograms(res["programs"], res["profiles"], metaclusters)

Layout

• pyscalop/ — importable package
• analysis/ — analysis scripts (source functions, consume data, emit results/plots)
• data/ — important intermediate data
• tests/

Status

v0.1 — core modules ported: score, dea, programs, metaprograms, utils, plot.