CHANGELOG.md

Changelog

All notable changes to this project will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

[v1.4.0]

Added

• BED files for the VNTR regions in hg19/hg38 from Severus.
  • The appropriate file will be automatically selected for the appropriate genome, unless a user provides a custom bed with --tr_bed.
• IGV configuration supporting VCF files from --snv and --sv.
• Aligned BAM to the output, when re-alignment is required.

Changed

• Tweaked parameters for Severus to refine SV calling.
• More informative log when a BAM called with an invalid basecaller model is provided.
• Failures of processes involved in differentially modified loci and regions detection will not cause workflow to fail.
• Updated modkit to v0.3.3.
• Reconciled workflow _ingress.nf from wf-human-variation v2.4.0 and wf-template v5.2.6.

Fixed

• -resume failing for some snv processes.
• Excessive memory usage for sample_probs process when using --mod leading to exit code 137.
• Erroneous handling of inputs for the joint report.
• Missing re-aligned XAM files in the output directory.

[v1.3.1]

Added

• --override_basecaller_cfg parameter allows users to provide a basecall configuration name in cases where automatic basecall model detection fails.
• --diff_mod option to allow users to turn off differential modified loci (DMR) and regions (DMR) analysis with DSS by setting --diff_mod false.

Changed

• Updated modkit to v0.3.0.
• Reconciled workflow _ingress.nf from wf-human-variation v2.3.1.

Fixed

• --snv crashing with --include_all_ctgs true.
• --snv always referring to model in --override_basecaller_cfg when deciding whether to call Indels.
• Automated basecaller detection not finding a basecaller model.

Removed

• --basecaller_cfg as the workflow now automatically detects the basecaller model from the input data.

[v1.3.0]

Added

• Tumor-only mode for the base workflow, small variant calling with ClairS-TO, modified base aggregation and somatic SV calling.

Fixed

• Parts of the documentation still referring to nanomonsv.

Changed

• If available basecaller_cfg will be inferred from the basecall_model DS key of input read groups.
  • Providing --basecaller_cfg will not be required if basecall_model is present in the DS tag of the read groups of the input BAM.
  • basecaller_cfg will be ignored if a basecall_model is found in the input BAM.
  • The workflow will fail if the tumor and normal BAM files have not been called with the same basecall_model.
• Updated to Severus v1.1.

Fixed

• Workflow crashing when the input BED file has overlapping intervals.
• Returning error in annotate_sv when END position smaller than POS.
• Workflow not starting in nextflow v24.04.

[v1.2.2]

Changed

• Update to ClairS v0.2.0.
  • The workflow now uses normal heterozygote sites to haplotag reads.
  • This behaviour can be changed using --use_normal_hets_for_phasing and --use_tumor_hets_for_phasing.
  • Moreover, it uses the indels for the phasing as default; this can be changed with --use_het_indels_for_phasing.
  • The workflow now uses longphase to haplotag reads; this can be changed with --use_longphase_haplotag.
  • The workflow now accepts a --liquid_tumor option, enabling presets and, where available, models specific for liquid tumors.
• Update to Clair3 v1.0.8.
  • Added --clair3_base_err and --clair3_gq_bin_size options.
• modkit now runs by contig.
• modkit bedMethyl are now in the top level output directory.

[v1.2.1]

Added

• A report with name [sample name].wf-somatic-variation-report.html, linking the individual detailed reports.

Changed

• Use ezcharts SeqCompare to in QC report.
• Memory usage of alignment report reduced by using histograms.
• Retry process when clairs.py predict crashes with error 134.

[v1.2.0]

Added

• Support for input folders of BAM files for --bam_tumor and --bam_normal (instead of only allowing single BAM files).

Changed

• ClinVar version in SnpEff container updated to version 20240307
• Update to Clair3 v1.0.7.
• Update to modkit v0.2.6.
• Improved modkit runtime by increasing the default interval size.
• Increased minimum CPU requirement for the workflow to 16.
• bedMethyl output files now follow the pattern {{ alias }}.wf-somatic-mods.{{ type }}.bedmethyl.gz.
• Structural variant (SV) calling is now performed with Severus (v0.1.2).
• ARM-compatible base workflow and modified base calling.
• minimap2 alignments will be in BAM format when --sv is set.

Fixed

• Force minimap2 to clean up memory more aggressively. Empirically this reduces peak-memory use over the course of execution.
• Workflow occasionally repeating QC analyses when resuming, even if successful.
• Alignment report script using too much memory.
• HAC models not recognised as valid.
• Spurious bamstats crashes with implausible alignment information.

Removed

• CRAM as supported input format.
• Reference genome and its indexes from the output directory.
• Insert classification and support for mismatching panel of control.
• Options --qv, --classify_insert, --min_ref_support, --genotype_sv and --control_panel, as these are no longer used by the workflow.

[v1.1.0]

Changed

• Updated ClairS to v0.1.7, with the new dorado 4KHz/5KHz HAC models.
• Several performance improvements which should noticeably reduce the running time of the workflow
• makeQCreport allows for one retry to prevent the workflow failing on report generation

[v1.0.0]

Added

• Tumor-only mode for the base workflow and modified base aggregation.
• --control_panel option to provide a non-matching control panel produced with nanomonsv merge_control.
• Memory directives for every process.

Changed

• Run ClairS haplotype_filter by contig.
• --sv will no longer emit genotypes by default, and only save the same fields from nanomonsv.
  • Users can still request a genotype using --genotype_sv.
  • --min_ref_support defines the minimum number of REF-supporting reads to call a heterozygote site.

Fixed

• --snv calling genome-wide variants when --bed is specified.
• snv:makeReport crashing when --annotation false.

Removed

• --annotation_threads removed as snpEff v5 uses only one thread.

[v0.5.2]

Changed

• New documentation
• Updated ClairS to v0.1.6
• ClinVar annotation of SVs has been temporarily removed due to not being correctly incorporated. SnpEff annotations are still produced as part of the final SV VCF.

Fixed

• rVersion retried also upon success

Removed

• Default local executor CPU and RAM limits

[v0.5.1]

Added

• List of reads supporting the SV events is now emitted in {params.output}/{params.sample_name}/sv/txt

Fixed

• Running in --sv mode does not resume properly.
• somatic_sv:report process failing due to name collisions when running with --annotation false.
• Modifed base calling report showing overlapping lines in the DMR plot.

[v0.5.0]

Added

• Automated annotation of SNVs, small indels and SVs.
• Option to skip germline calling and variant phasing with --germline false.
• Workflow can now emit germline GVCFs for tumor/normal samples with --GVCF.
• Option --normal_vcf to provide a pre-computed normal VCF file.
• Add genotyping and hybrid mode.
• The workflow saves the haplotagged cram files if --germline true.

Changed

• Updated modkit to v0.1.13, ClairS to v0.1.5, Clair3 to v1.0.4, and added support for 4KHz and 5KHz dorado models.
• Runs ClairS haplotype_filter on the indels in addition to the SNVs.
• SV VCF now report a single sample with Tumor/Normal formats collected, rather than two distinct samples
  • This facilitates merging multiple samples thanks to unambiguous sample labelling
  • The original VCFs generated by nanomonSV are now saved in ${outdir}/${sample_name}/sv/vcf
• More consistent output naming formatting.
• When --bed and --<tumor|normal>_min_coverage are specified, the workflow will process the regions with coverage above the threshold
  • The filtering will retain the union (bedtools merge -d 0) of the tumor and normal filtered regions

Fixed

• snpEff crashing due to running out of memory.
• Error in annotate_sv when END position smaller than POS

[v0.4.0]

Added

• Automated annotation of SNVs and small indels.
  • Disable with --annotation false
• ARM-compatible base workflow and modified base calling
• Option --qv, to specify the expected quality value for nanomonsv

Changed

• Input options and Output options have been combined in the Main options category
• Updated DSS to v2.38.0
• Updated modkit to v0.1.12
• Updated nanomonsv to v0.7.1

Fixed

• The workflow is now mostly species agnostic (with the exception of --annotation and --classify_insert)

[v0.3.0]

Added

• Add modified base calling with --mod

Changed

• Updated to nanomonsv v0.6.0.
• Refine change type counts and plot; file named [SAMPLE]_spectrum.csv renamed to [SAMPLE]_changes.csv.

[v0.2.0]

Added

• Coverage plots to alignment stats report

Changed

• Improved documentation and new structure of the output sub-directories
• Variant allele frequency representation is now a scatterplot showing the relationship between normal and tumor VAF
• Created a subworkflow to call somatic SV (somatic_sv in workflows/wf-somatic-sv.nf)
• Add nanomonsv soft filtering SV when providing a bed file specifying the tandem repeat with --tr_bed
• Add nanomonsv insert classification with --classify_insert, to add RepeatMasker annotation to the SVs
• Add report_sv, that generates a report of the SV detected
• Enum choices are enumerated in the --help output
• Enum choices are enumerated as part of the error message when a user has selected an invalid choice
• Bumped minimum required Nextflow version to 22.10.8

Fixed

• Replaced --threads option in fastqingress with hardcoded values to remove warning about undefined param.threads
• Depth plot with overlapping chromosomal coverage

[v0.1.1]

Fixed

• Demo data
• Fix occasional crash when candidate nested tuple is found

Added

• Configuration for running demo data in AWS
• Fast mode

Changed

• Updated to ClairS v0.1.1

[v0.1.0]

Changed

• Initialised wf-somatic-variation from wf-template
• Implemented wf-somatic-snp module, that runs ClairS in a highly parallelised way (v0.1.0).
• Implemented some accessory modules to visualise the results from ClairS (mutation counts, variant allele frequency).
• Implemented report of alignment statistics
• Update reporting script to use ezcharts
• Customizable thread count for haplotype filtering stage
• Mosdepth per-base statistics are now optional, and are not shown in the report

Fixed

• Workflow crashing when predicting in regions without variants
• Workflow crashing when concatenating SNP and Indel VCF files
• Workflow interrupting when an empty Indel VCF file is generated
• Extremely slow reporting of alignment statistics