Skip to content

Latest commit

 

History

History
66 lines (33 loc) · 1.96 KB

File metadata and controls

66 lines (33 loc) · 1.96 KB

task_label_projection devel

NEW FUNCTIONALITY

  • Add new method: CellMapper, which is a k-NN based approach to map cells across representations and can be used for label projection. Two versions are included here, one based on PCA or CCA embeddings (linear) and one based on an scvi embedding (scvi) (PR #22)

MAJOR CHANGES

  • Update the image version from :1.0.0 to :1 for openproblems/base_* images (PR #21)

MINOR CHANGES

  • Update scPRINT to better handle large datasets, including a new default model (PR #20)

  • Update Geneformer for new model versions (PR #24)

  • Update scPRINT to use latest stable version (2.3.5) with compatible dependencies and fixed installation method (PR #25)

  • Update Viash version to v0.9.4 (PR #23)

task_label_projection 2.0.0

A major update to the OpenProblems framework, switching from a Python-based framework to a Viash + Nextflow-based framework. This update features the same concepts as the previous version, but with a new implementation that is more flexible, scalable, and maintainable.

Migration

  • Added expected input/output interfaces in src/api and document them in README.md.

  • Store common resources used across tasks in a git submodule common.

  • Methods, metrics, workflows and other components are implemented as Viash components with a per-component Docker image.

New functionality

  • Switched to larger datasets derived from CELLxGENE.

  • Added scGPT zero shot (PR #2).

  • Added scGPT fine-tuned (PR #3).

  • Added SCimilarity (PR #4).

  • Added UCE method (PR #6).

  • Added geneformer (PR #7, #16).

  • Added scPRINT (PR #8).

Major changes

  • Updated the task API (PR #9).

Bug fixes

  • Convert to dgCMatrix in SingleR (PR #5).

  • Multiple fixes prior to release (PR #11, #13, #14, #15, #17).

Documentation

  • Update README (PR #10).

task_label_projection 1.0.0

This version can be found here.