SINCLAIR 0.4.0

• All nextflow processes now use containers rather than environment modules for dependencies. (#95, @kelly-sovacool)
• Creates cache for celldex reference downloads in working directory (#203, @wong-nw)
• Checks for joined RNA "layers" in Seurat v5 object (#224, @wong-nw)
• Support rliger versions < 2.0 and >= 2.0. (#228, @wong-nw)

SINCLAIR 0.3.7

• Improvements to R Markdown reports: (#216, @kelly-sovacool)
  • Link to function documentation websites.
  • Use code folding to include code in output, but set it to 'hide' by default.
  • Hide R messages by default.
• Updates for compatibility with Nextflow v26. (#233, #234, @kelly-sovacool)
• Silence warning in Nextflow onComplete handler. (#239, @kelly-sovacool)

SINCLAIR 0.3.6

• Set main job memory to 2 GB for slurm mode. (#205, @kelly-sovacool)

SINCLAIR 0.3.5

• Fix samplesheet check to allow multi-digit sample IDs. (#189, @kelly-sovacool)
• New parameter save_cellranger_extra_files to control whether to save extra cellranger output files (bam, bai, cloupe) in addition to h5 files. Default is false to save space. (#195, @kelly-sovacool)
• New largemem profile for memory-intensive processes. (#196, @kelly-sovacool)

Documentation updates

• Major improvements to the docs website. (#160, @bianjh-cloud)
• Fix rendering of bulleted lists in docs website. (#164, @kelly-sovacool)
• Documentation for test data outputs updated. (#174, @wong-nw)
• Minor documentation improvements (#175, @wong-nw)
• Update test-info.md to reflect change in directory naming (#190, @wong-nw)

Other minor changes

• Re-ordering of batch correction report. (#174 @wong-nw)
• Rename 'output' directory to 'results' for the nextflow workflow. (#186, @kelly-sovacool)

SINCLAIR 0.3.4

• Improve error messages when checking the sample sheet and contrast sheet. (#181, #182, @kelly-sovacool)

SINCLAIR 0.3.3

• Improve help message for sinclair run. (#153, @kelly-sovacool)
  • the default --mode is now slurm.
• The run_cellranger parameter is now a boolean set to true by default. (#155, @kelly-sovacool)
  • Set this to true if the input is fastq files or false if filtered h5 files.
• New profiles:
  • test_h5 for starting from h5 files. (#155, @kelly-sovacool)
  • debug, which sets process.scratch to false so intermediate files will be in the workDir. (#155, @kelly-sovacool)
• Fix file paths for environments and repos on biowulf. (#162, @wong-nw)

SINCLAIR 0.3.2

• Improve error messages in SAMPLESHEET_CHECK processes. (#143, @kelly-sovacool)
• SINCLAIR now depends on ccbr_tools v0.4 for updated jobby & spooker utilities. (#147, @kelly-sovacool)

SINCLAIR 0.3.1

• Minor documentation improvements. (#135, @kelly-sovacool)
• SINCLAIR is now archived in Zenodo and can be cited with the DOI 10.5281/zenodo.15283503. (#136, @kelly-sovacool)
• Fix bug where nextflow schema was not included in the installation. (#138, @kelly-sovacool)

SINCLAIR 0.3.0

• Breaking change: GEX is now the default workflow. The -entry argument is no longer used. (#129, @kelly-sovacool)

New features

• Allows users to determine what variables to regress out. (#55, @slsevilla)
• Overhaul the CLI to use python rather than bash, which introduces breaking changes (#61, @kelly-sovacool).
  • Create a script (bin/sinclair) to provide an interface to the CLI that works out-of-the-box without the need to install the python package with pip. (#80, @kelly-sovacool)
• Use nextflow run -resume by default, or turn it off with sinclair run --forceall. (#110, @kelly-sovacool)
• Add --output argument for sinclair init and sinclair run. (#110, @kelly-sovacool)
  • If not provided, commands are run in the current working directory.
  • This is equivalent to the nextflow $launchDir constant.
• Set the publish_dir_mode nextflow option to link by default. (#110, @kelly-sovacool)
• Set the process.cache nextflow option to deep by default rather than lenient on biowulf. (#110, @kelly-sovacool)
• Before launching the pipeline run:
  • The nextflow preview is printed. (#117, @kelly-sovacool)
  • The nextflow parameters are validated. (#127, @kelly-sovacool)

Bug fixes

• Fix biowulf module syntax in conf/modules.config. (#81, @epehrsson)
• Fix filtering thresholds and use filtered object for downstream steps in SEURAT_PROCESS. (#81, @epehrsson)
• Fix seurat object and group assignment in SEURAT_MERGE. (#81, @epehrsson)
• Correctly map h5 files from cellranger with their input fastq files during preprocessing. (#82, @kelly-sovacool)
• Fix the $SLURM_JOB_ID variable name for biowulf. (#82, @kelly-sovacool)
• Fix file paths for test dataset. (#82, @kelly-sovacool)
• Use same number of PCs for merged object clustering as for integration. (#85, @epehrsson)
• Add LIGER UMAP to integration report. (#85, @epehrsson)
• Set all default parameters in nextflow.config. (#85, @epehrsson)
  • Previously, some parameters were set in conf/process_params.config, but we found this confusing, so we consolidated them to the main nextflow.config file.
• Allow sample IDs to contain hyphens. (#94, @wong-nw)
• Disable SCVI batch correction. (#109, @wong-nw)
  • This feature is on hold until a later release.
• LIGER now runs with 50 PCs by default instead of 20 (#109, @wong-nw)
• Output all R Markdown documents as HTML rather than PDF. (#112, @kelly-sovacool)
• Make sure values in the contrasts sheet are treated as strings. (#133, @kelly-sovacool)

Documentation improvements

• The docs website now has a drop-down menu to switch between different versions. (#103, @kelly-sovacool)
• Fix broken image link. (#108, @wong-nw)
• Now using the readthedocs theme. (#111, @kelly-sovacool)
• Update to include quickstart and clarify existing documentation (#122, @wong-nw)

v0.2.0

• functioning GEX workflow, with modules pulled from Biowulf