clustering-degs

name	clustering-degs
description	Find differentially expressed genes (markers) for clusters using wilcoxon, t-test, or logistic regression. Includes consensus markers and resolution scanning.
allowed-tools	Read, Bash
compatibility	scRNA-seq, snRNA-seq

Clustering Differential Expression

When to Use

After clustering to identify marker genes for each cluster, or before annotation to refine cluster resolution.

Functions

find_markers(adata, groupby, method="wilcoxon", top_n=20) -> dict

Finds DEGs per cluster.

Parameters:

• method: "wilcoxon" (default), "t-test", or "logreg"
• top_n: number of top markers per cluster

Returns dict keyed by cluster ID with list of gene names.

consensus_markers(adata, groupby, methods=("wilcoxon", "t-test"), min_methods=2, top_n=20) -> dict

Finds markers that appear in results from multiple statistical methods.

Parameters:

• methods: tuple of methods to compare
• min_methods: minimum number of methods that must agree (default 2)

Returns dict of consensus markers per cluster.

resolution_scan(adata, resolutions, cluster_key="leiden", quality_fn="silhouette") -> dict

Scans multiple resolution values and returns quality scores.

Returns {resolution: score} dict and recommends best resolution.