Skip to content

Changelog

[2026-10-13] — v4.0.0

Overview

G4X-helpers v4 is a major architectural release designed to support the latest G4X-data v4 format and the new zarr based G4X-viewer input. This release reorganizes the Python API, processing modules, CLI, schema validation, and G4X-viewer generation. It introduces a more consistent workflow for processing, validating, migrating, and modifying G4X datasets.

Highlights

  • continued support to migrate legacy data to latest v4 format
  • now handles single-cell processing steps after redemux and resegment operations
  • begin phase-out of jp2 files (will be replaced with ome.tiff during migration)
  • new functions to inspect and edit G4X-viewer input
  • added optional GPU acceleration through the gpu dependency extra.
  • added region-of-interest support for processing cropped datasets.
  • expanded support for transcript-only, protein-only, and combined assay data.

CLI

  • Reorganized the CLI around five supported commands:
  • redemux
  • resegment
  • migrate
  • validate
  • viewer
  • Added migrate --check for checking migration compatibility without modifying data.
  • Added migrate --roi for migrating a selected region.
  • Added branch-based output handling for redemux and resegment.
  • Added --no-downstream for skipping downstream processing.
  • Added validate --raw-only for validating only the inputs required by G4X-helpers.
  • Added viewer metadata import and export commands for images, cells, and transcripts.
  • Improved CLI help messages, progress reporting, error handling, and version reporting.

Data processing

  • Refactored demultiplexing, aggregation, and single-cell processing into independent modules.
  • Added configurable demultiplexing parameters and batch handling.
  • Added RNA and protein correlation analysis.
  • Added assay-aware image, transcript, and cell output generation.
  • Improved handling of missing bead masks, empty transcript data, failed clustering, and transcript-only runs.

Validation and migration

  • Replaced the legacy schema implementation with dedicated file, directory, and dataset validators.
  • Added more detailed validation reporting for raw and processed G4X-data.
  • Reworked legacy-data migration around dedicated migrator classes.
  • Improved samplesheet, manifest, transcript-table, protein-directory, and sample.g4x validation.
  • Added migration support for cropped regions and updated G4X-viewer layouts.

G4X-viewer

  • New G4X-viewer Zarr creation for images, transcripts, cell masks, and metadata.
  • Added metadata import and export for existing viewer stores.
  • Added support for multiple segmentations and assay-aware channel ordering.
  • Improved image conversion, tiling, chunking, and Zarr compatibility.
  • Improved transcript visualization, polygon generation, and color controls.

Reliability and release infrastructure

  • Expanded CLI, processing, smoke, transcript-only, and protein-only tests.
  • Added strict documentation, package, wheel, and multi-architecture Docker preflight checks.
  • Added AMD64 and ARM64 container validation.
  • Refreshed dependencies to address reported security vulnerabilities.
  • Removed development and documentation dependencies from production container images.
  • Updated GitHub Actions to use Node.js 24-compatible action versions.

[2026-04-13] — v3.0.2

  • fix: load adata from correct location

[2026-04-03] — v3.0.1

  • update changelog
  • remove "staging" as release-branch in uv-ship

[2026-04-03] — v3.0.0

  • feat: compatibility with "cytoplasmic" image label
  • fix: ensure new_bin feature handles init_bin logic
  • fix: resegment not updating bin file
  • fix: ensure new gene names propagate to sc-out during redemux
  • fix: cannot save adata object

[2025-12-30] — v2.1.2

  • Fix/auto revert migrate (#68)

[2025-12-29] — v2.1.1

  • fix: migration tripping over file updates despite fail

[2025-12-29] — v2.1.0

  • feat: "validate" function for detailed schema output validation (#67)
  • fix: re-demux on tx-only runs (#66)

[2025-12-16] — v2.0.3

  • fix: prevent validation fail for tx-only runs (#64)
  • fix: code scanning alert no. 3: Workflow does not contain permissions (#62)
  • pkg: incorporate test data from S3 for full feature tests (#61)
  • docs: layout updates and cleaner changelog format (#63)

[2025-12-09] — v2.0.2

  • docs: update to match global layout
  • docs: set up section indices correctly
  • docs: correct readme reference
  • pkg: clean up changelog
  • pkg: pre-commit

[2025-12-04] — v2.0.1

  • fix: docker release auth issue

[2025-12-04] — v2.0.0

Overview:

  • New "migrate" function to update older datasets to be compatible with the latest G4X-viewer and G4X-helpers versions.
  • Added schema validation to ensure data integrity and compatibility.
  • Implementation of an initial test framework for the package.
  • Re-worked CLI-api to be more modular and easier to extend.

Changes:

  • new "in_place" output flag replaces "out_dir" and ensures predictable output behavior.
  • standardized log location
  • new logging formatter
  • refactor "workflows" into modules
  • adapt G4Xoutput to latest schema
  • refactored stream_features method to redemux module
  • new write_csv_gz method
  • add bead_mask loading to G4Xoutput
  • removed deprecated G4X-viewer schema
  • new dependency: pathschema
  • updates dependencies
  • update docs content and match structure to main site
  • update changelog

[2025-12-03] — v1.0.2

  • fix: update uv.lock

[2025-12-01] — v1.0.1

  • fix: get_shape fast via glymur

[2025-11-13] — v1.0.0

  • docs: add partials for shared options
  • pkg: add rich-codex as docs dependency
  • fix: load_segmentation method now allows custom keys
  • feat: redemux parses transcript panel for input flexibility
  • docs: update for new CLI
  • feat: numpy version agnostic npzGetshape
  • pkg: repr update
  • pkg: implement new cli and re-factor
  • pkg: set python 3.12 as default version

[2025-10-01] — v0.5.2

  • fix: missing tests in publish workflow

[2025-10-01] — v0.5.1

  • pkg: added PyPI publish workflow
  • fix: README logo link

[2025-10-01] — v0.5.0

  • docs: formatting
  • docs: spell check
  • docs: integrate redemux docs
  • docs: Docs restructuring (#33)

  • docs: updated _core docs

  • feat: Redemux tool (#30)

  • feat: progress bar

  • fix: G4Xoutput out_dir now defaults to cwd instead of run_base

  • fix: enforce cluster color as hex-codes
  • fix: auto-conversion of clusters to 'categorical' when load_clustering=true
  • fix: change all segmentation_cell_id to cell_id (#34)

  • fix: cellid_key - bugfix

  • pkg: updated ruff

  • pkg: relaxing glymur dependency
  • pkg: relaxed protobuf requirement for internal alignment
  • pkg: opening python requirements to include 3.11
  • pkg: removed bump-my-version and replacing with uv-ship

[2025-08-12] — v0.4.14

Fixes:

  • tar_viewer will now exit leaving the source folder unchanged

Improvements:

  • tar_viewer now requires an out_path

[2025-08-11] — v0.4.13

Docs:

  • changelog now included in docs
  • url updates
  • typos

Package:

  • deployment of multi-arch packages
  • replaced dependency ray with multiprocessing
  • set required uv version

[2025-07-25] — v0.4.12

Docs:

  • Added Documentation for G4X-helpers and G4X-output

Package:

  • Implemented bump-my-version for handling version updates and tagging
  • Added pre-commit hooks for code quality checks
  • Ruff for linting and formatting (invoked via pre-commit)
  • Github actions for automated package builds and docs deployment

Improvements:

  • Added npz util to speed up G4Xoutput() initialization

Fixes:

  • Incorrect gzip compression on some output csvs

Housekeeping:

  • Updated and trimmed dependencies
  • Cleaned up .gitignore file
  • Cleaned up README.md

[unreleased changes]

  • Release preparation
  • Add Dockerfile
  • Update dependency information
  • Add tar_viewer tool to tar up a G4X-viewer folder for the single-file upload option.
  • Add new_bin tool to more quickly generate a new bin file
  • Bug fixes for MVP functionality
  • Add CLI tools for re-segmentation and updating bin files with clustering/embedding information