CHAMPAGNE 0.5.3

• CHAMPAGNE now requires nextflow version 25.10.0 or later. (#303, @kelly-sovacool)
• Increase memory for: (#298, @kelly-sovacool)
  • main slurm job
  • diffbind
  • deeptools plot_heatmap

CHAMPAGNE 0.5.2

• Calculate spike-in scaling factors on deduplicated reads after aligning to the spike-in genome. (#289, @kelly-sovacool)
• Improve the spike-in scaling factor table to include the antibody used for each sample in the MultiQC report. (#290, @kelly-sovacool)
• Include an example MultiQC report in the documentation. (#291, @kelly-sovacool)

CHAMPAGNE 0.5.1

• Minor documentation improvements. (#273, @kelly-sovacool)
• Set --blacklist to a custom bed or fasta file to override the default blacklist used by a built-in --genome. (#278, @kelly-sovacool)
  • Also add support for the hg19 genome on biowulf.
• Fix spike-in normalization: (#279, @kelly-sovacool)
  • Do not include inputs in the scale factor calculation.
  • Group samples by antibody for scale factor calculation.
  • Create separate parameters for controlling the deeptools normalization method for samples and inputs.

CHAMPAGNE 0.5.0

• CHAMPAGNE now depends on ccbr_tools v0.4 for updated jobby & spooker utilities. (#247, @kelly-sovacool)
• Reorganized output directory structure. (#256, @kelly-sovacool)
• Nextflow parameters are now validated at the start of the workflow. (#257, @kelly-sovacool)

New features

• Contrasts are now specified as a TSV file instead of YAML. (#224, @kelly-sovacool)
  • See the example contrast sheets in assets/.
• New consensus peak method from Corces et al. (doi:10.1126/science.aav1898). (#225, #246, @kelly-sovacool)
• Normalization with a spike-in genome is now supported. (#259, @kelly-sovacool)
  • Set params.spike_genome to dmelr6.32 or ecoli_k12
  • The normalization method can be set with params.spike_norm_method to delorenzi (default) or guenther. (#267, @kelly-sovacool)
• CLI improvements
  • Use nextflow run -resume by default, or turn it off with champagne run --forceall. (#224, @kelly-sovacool)
  • Add --output argument for champagne init and champagne run. (#232, #233, @kelly-sovacool)
    • This is equivalent to the nextflow launchDir constant.
  • --mode now has slurm as the default. (#249, @kelly-sovacool)
• Improved nextflow options:
  • Set publish_dir_mode to link by default.
  • Set process.cache to deep by default rather than lenient. (#224, @kelly-sovacool)
  • Enable the nextflow timeline & trace reports by default. (#226, @kelly-sovacool)
  • Output BioCompute objects by default using the nf-prov plugin. (#260, @kelly-sovacool)

Bug fixes

• Refactor checks for the sample sheet & contrast sheet to prevent unnecessary re-running. (#224, @kelly-sovacool)
• Fix a file name clash during input pooling. (#224, @kelly-sovacool)
• Fix bug in MEME AME process that caused it not to run on all samples. (#234, @kelly-sovacool)
  • Also correct the motif rank calculation. (#234, @kopardev)
• Ensure inputs are pooled correctly for paired end reads. (#250, @kelly-sovacool)
• Fix chipseeker filename collision. (#252, @kelly-sovacool)
• Allow fraction-in-peaks plot to run when there is only 1 sample. (#265, @kelly-sovacool)

Documentation

• Now using the readthedocs theme for the docs website. (#236, @kelly-sovacool)
• Improved help message for champagne run. (#249, @kelly-sovacool)
• Major improvements to fill out the documentation website. (#262, #263, @kelly-sovacool)

CHAMPAGNE 0.4.1

• The CHAMPAGNE nextflow workflow now has a version entry in nextflow.config, in compliance with nf-core. (#213, @kelly-sovacool)
• Pool input (control) reads of the same sample name by default. Any inputs that should not be pooled must have different sample names in the samplesheet. (#214, @kelly-sovacool)
• Add histone samples to the test_human dataset. (#215, @kelly-sovacool)

CHAMPAGNE 0.4.0

New features

• Create a script (bin/champagne) to provide an interface to the champagne CLI that works out-of-the-box without the need to install the python package with pip. (#180, @kelly-sovacool)
  • However, any dependencies not in the Python Standard Library must be installed for this to work. See the dependencies list in pyproject.toml.
• Allow additional columns in the sample sheet beyond the minimum required header. (#176, @kelly-sovacool)
• Add a workflow entry point to download fastq files from SRA. (#176, @kelly-sovacool)
• Add test_human profile with chipseq data from ENCODE. (#176, @kelly-sovacool)

Bug fixes

• Fix configuration files for compatibility with using the GitHub repo as the source. (#173, @kelly-sovacool)
  • These equivalent commands now work:

    nextflow run CCBR/CHAMPAGNE
    champagne run --main CCBR/CHAMPAGNE

• Allow multiple samples to use the same input. (#176, @kelly-sovacool)
• In the biowulf config profile, switch variable $SLURM_JOBID to $SLURM_JOB_ID. (@kelly-sovacool)
• Increase resource allocations for chipseeker and deeptools. (#192, @slsevilla)
• Check the validity of the contrastsheet earlier on in the workflow. (#192, @slsevilla; #200, @kelly-sovacool)
• Fix bug where manorm was using R1 twice instead of R1 and R2. (#206, @kelly-sovacool)

Misc

• Change the peak widths histogram type from overlay to stack. (#176, @kelly-sovacool)
• Documentation improvements. (#192, @slsevilla)

CHAMPAGNE 0.3.0

New features

• Find motifs in the genome with Homer. (#142)
• Run motif enrichment analysis with MEME. (#142)
• Annotate peaks with chipseeker. (#142,#147,#157)
• Add preseq complexity curve and fastq screen to multiqc report. (#147)
• Support multiple replicates per sample and call consensus peaks on replicates. (#129)
  • Optionally normalize p-values with the CCBR/consensus_peaks subworkflow.
• Implement differential peak calling. (#158)
  • Optionally specify contrasts via a YAML file. If no file is specified, differential analysis is not performed.
  • If any sample has only one replicate, run MAnorm, otherwise run diffbind.
• Print the recommended citation in bibtex format with champagne --citation. (#153)
  • CHAMPAGNE is also now archived in Zenodo with DOI 10.5281/zenodo.10516078.
• The docs website now has a dropdown menu to select which version to view. The latest release is shown by default. (#170)

Bug fixes

• Fix deepTools plots (#144):
  • Per sample fingerprint plots instead of per replicate.
  • Input normalized profile plots.
  • Protein-coding-only versions of plots.
  • Ensure sample IDs are sorted. (#150)
• Fix a bug where the wrong SICER output file was used for downstream analyses. (#155)
• Fix CLI profile on machines other than biowulf & FRCE. (#168)
• Fix broken bold styling in documentation website. (#53)

CHAMPAGNE 0.2.2

• Fix permissions issues in the CLI. (#167)

CHAMPAGNE 0.2.1

• Fixed a bug in QC stats that mixed up the statistics for different samples. (#125)
• Fixed a bug in the CLI that added the -profile to the nextflow command even if it wasn't needed (#125).
• Report read counts between blacklist & filtering steps in the QC table. (#125)
• Run spooker on workflow completion (#126).

Full Changelog: v0.2.0...v0.2.1

CHAMPAGNE 0.2.0

New features

• Implemented peak calling with sicer2, macs2, and gem. (#52)
• Added parameter options to skip QC, input normalization, and/or peak calling steps. (#72)
• Calculate and plot QC metrics for called peaks:
  • Fraction in Peaks (FRiP) (#89)
  • Jaccard index (#92)
  • Histogram of peak widths (#92)
• Added support for paired-end reads. (#105)
• Added an option to use a custom reference from a genome fasta, gtf, and blacklist file. (#105)
• Champagne CLI: (#112)
  • New --mode option for champagne run to execute the workflow locally ('local') or submit it as a slurm job ('slurm').
  • Option to override the path to the champagne main.nf file or specify the github repo (CCBR/CHAMPAGNE) instead.

    # use the default path
    champagne run ...
    # override the path
    champagne run path/to/champagne/main.nf
    # use a revision from github instead
    champagne run CCBR/CHAMPAGNE -r v0.1.0

Bug fixes

• CLI:
  • Error when biowulf-specific environment variables are not defined. (#54)
  • The host is now correctly detected as biowulf via scontrol. (#75)
• Containers:
  • Containers are now specified in process definitions instead of withName/withLabel for better control. (#69)
    • Shared containers are specified as parameters in the config file conf/containers.config.
  • No longer use --mount type=bind or --volume for making directories available to processes in containers. Instead, use Nextflow's Channel.fromPath constructor with type: 'dir'. (#71)

API-breaking changes

• An error is thrown when a required input file doesn't exist. (#71)
  • Previously, the workflow quietly didn't run the process(es) that required the missing file.
• Renamed champagne config to champagne init to avoid clashing with nextflow config. (#112)

Full Changelog: v0.1.0...v0.2.0