feat: FilterMarkers

[maxkarlsson]

Description

Added FilterMarkers which can be used to get a marker list that fulfills some criteria in a Seurat object. I am uncertain about the naming here, as the camel case FilterMarkers implies that it can be pipable Seurat-style, but it can't since the output is a only list or vector.

Added

• FilterMarkers to keep markers whose CPM exceeds an isotype-relative
  ratio and/or an absolute abundance threshold in at least
  min_cell_fraction of cells.

Type of change

• New feature (non-breaking change which adds functionality).

How Has This Been Tested?

Tests are added.

PR checklist:

• My changes generate no new warnings.
• I have added tests that prove my fix is effective or that my feature works.
• I have documented any significant changes to the code in CHANGELOG.md

---

Note

Low Risk
Additive exported API with validation and tests; read-only on Seurat data and no changes to auth, I/O, or core assay pipelines.

Overview
Adds FilterMarkers (v 0.20.2) so users can derive a marker panel from a Seurat object’s counts layer using CPM-based positivity rules, alongside existing isotype helpers like isotype_pls.

A marker is kept when at least min_cell_fraction of cells are “positive,” where positivity uses isotype_ratio (vs. median isotype CPM), optional abundance_threshold, or the stricter combination of both. Isotype controls are always excluded. Optional group_column returns per-group results; return_stats returns a per-marker tibble instead of names only. The function does not subset or mutate the Seurat object—it returns a character vector, tibble, or named list.

Release notes, export/NAMESPACE, roxygen/man, spelling (CPM), and testthat coverage are included.

^{Reviewed by Cursor Bugbot for commit 7ffe604. Bugbot is set up for automated code reviews on this repo. Configure here.}

[Copilot]

🟢 Approval recommended

The feature is additive with tests and validation, and the only flagged issue is a minor roxygen documentation punctuation fix.

Pull request overview

Adds a new exported helper (FilterMarkers) to derive a marker panel from a Seurat object’s counts layer based on CPM positivity rules, plus a patch-level version bump and accompanying docs/tests.

Changes:

• Introduces FilterMarkers() with optional isotype-relative and/or absolute CPM thresholds, optional per-group operation, and optional stats output.
• Adds testthat coverage for default, threshold variants, stats output, grouping, and input validation errors.
• Updates package metadata/documentation for the new API and bumps version to 0.20.2 (README badge, DESCRIPTION, CHANGELOG, WORDLIST, NAMESPACE).

File summaries

File	Description
R/filter_markers.R	Implements the new FilterMarkers() API with validation, grouping support, and optional stats output.
tests/testthat/test-filter-markers.R	Adds tests covering expected kept markers, threshold modes, stats output, grouping behavior, and error cases.
NAMESPACE	Exports FilterMarkers.
man/FilterMarkers.Rd	Adds generated Rd documentation for FilterMarkers.
DESCRIPTION	Bumps version to 0.20.2 and adds filter_markers.R to Collate.
README.md	Updates beta release badge to v0.20.2.
CHANGELOG.md	Adds 0.20.2 entry documenting FilterMarkers.
inst/WORDLIST	Adds CPM to spelling whitelist.

Review details

Files not reviewed (1)

• man/FilterMarkers.Rd: Generated file

• Files reviewed: 7/8 changed files
• Comments generated: 1
• Review effort level: Lite

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.

[ludvigla]

Looks great! The only comment I have is about the function name. FilterMarkers sounds a bit like it will remove markers from the Seurat object. Maybe FindAbundantMarkers? Seurat has FindVariableFeatures for feature selection.