Skip to content

Latest commit

 

History

History
617 lines (470 loc) · 30.4 KB

HISTORY.md

File metadata and controls

617 lines (470 loc) · 30.4 KB
Raw

2.3.1 (31 Aug 2023)

2.3.0 - 2.3.1 diff

  • Oviraptor: explicitly use shipped FASTA (pr146, issue143).
  • MultiQC: remove historical gold_standard background refdata
  • MultiQC: bump to v1.14+umccrise.0.0.14
  • vcf2maf: use bioconda::vcf2maf-umccr
  • install.sh: use libmamba solver instead of mamba
  • conda: explicitly add toml dependency to ummcrise env

2.3.0 (13 May 2023)

2.2.1 - 2.3.0 diff

  • PCGR: major update to v1.4.1 (from v0.9)
    • #123
    • CPSR: major update to v1.0.1 (from v0.6)
    • complete refactor of PCGR/CPSR installation and wrapper
    • vcf_stuff: bump v0.6.2 -> 0.6.4 (to adjust for PCGR changes)
    • ngs_utils: 2.10.2 -> 2.11.0 (for cpsr genes ensembl update)
    • cacao: separate conda env
  • vcf2maf: use umccr conda fork of latest main branch
  • Re-enable BPI
  • Removed conda envs and support for neoantigens, microbiome, crossmap
  • Buildspec: ignore DVC integration test

2.2.1 (18 Oct 2022)

2.2.0 - 2.2.1 diff

  • Skip BPI completely. This release differs from v2.2.0 in that it completely skips BPI.
    • We have had one case running BPI v1.5 which actually timed out (instead of erroring out due to OutOfMemory). This release could therefore be used when we have samples struggling at the BPI stage.

2.2.0 (18 Oct 2022)

2.1.1 - 2.2.0 diff

  • SV filter: remove SR < PR BND requirement (issue114, issue100, pr115).
  • BPI: deal with OOM issue by skipping it altogether upon error (issue88, gpgr pr55).
    • if sample fails BPI, the SV filters get adjusted by removing the BPI_AF-specific filter. The SV tables in the cancer report will also not display any AF information.
  • Summarise SNV counts throughout workflow (pr115, issue92).
  • Pierian: Improve input prep (issue112, pr113, pr116).
    • If >50K SNVs, filter to gene BED, then sequentially remove noncoding intergenic/intronic until the threshold is achieved.
  • Bump gpgr 1.2.5 -> 1.4.1:
  • Bump vcf_stuff 0.6.0 -> 0.6.2:
    • for hypermutated samples (>500K SNVs) we now simply filter to only cancer gene regions (vcf_stuff pr5, then changed in commit 3e011b).
  • Bump ngs_utils 2.9.2 -> 2.10.2:
    • Adding the all-genes BED file for subsetting hypermutated samples (>500K SNVs) and for Pierian (when still >50K SNVs) (ngs_utils pr4).
    • Adding the cancer-genes BED file for subsetting hypermutated samples (>500K SNVs) (ngs_utils pr5).
  • Bump oviraptor 1.1.20 -> 1.2.0:
  • General cleanup of codebase, including:
    • peddy-related code + MultiQC files
    • MultiQC conda build recipes
    • ribbon-related code
    • legacy (4yr) deployment ansible/vagrant code
    • dockerfiles/ directory

2.1.1 (15 Aug 2022)

2.0.2 - 2.1.1 diff

  • General cleanup of codebase (remove GRIDSS, GPL wrappers, submodules, circos configs etc.).
  • PCGR:
  • Update vcf_stuff to the conda-based 0.6.0 version (instead of pipping the master directly) (commit3b4afd).
    • This version enables VEP in PCGR to annotate only those variants within coding regions when there are > 500K variants (hypermutated samples). The variants without VEP annotation get filtered out in PCGR.
  • Cancer report relocation to gpgr (pr105, issue96).
  • Increase max mem for BPI (pr107, issue88).
  • Add documentation for SNV, SV, and PCGR workflows (pr102, pr109).

2.0.2 (22 Apr 2022)

2.0.1 - 2.0.2 diff

  • Use FORMAT/SQ value as QUAL for DRAGEN somatic variants when generating BCFtools stats
  • Update MultiQC reference data to align with above change
  • Remove DRAGEN germline fragment length histogram file from MultiQC reference set
  • Expand MultiQC template sample ordering and highlighting
  • Explicitly provide sample name and read group to MultiQC DRAGEN mapping metrics processor
  • Use NGS Utils 2.9.2 and umccr/MultiQC 1.13dev+umccrise.0.0.1 in conda environments
  • Improve DRAGEN sample identifier checks
  • Pin conda and mamba versions to avoid issues creating conda environments with newer versions

2.0.1 (15 Feb 2022)

2.0.0 - 2.0.1 diff

  • Handle multiple '@RG SM' BAM header fields (#83)

2.0.0 (30 Nov 2021)

1.2.4 - 2.0.0 diff

  • Overhaul automatic detection for DRAGEN inputs
  • Allow inputs to be provided via named arguments
  • Ignore timestamp for input BAMs and associated indexes
  • Enforce restriction of single bcbio batch or DRAGEN run
  • Prevent execution with bcbio data as input
  • Remove downsampled bcbio test from buildspec.yml
  • Run BCFtools, SAMtools stats appropriately for non-bcbio input
  • Update DRAGEN reference QC files
  • Upgrade to NGS Utils 2.9.1 (includes updated DRAGEN support)
  • Upgrade to MultQC 1.12dev+umccrise.0.0.2 (includes umccrise-specific patches)
  • Relabel QC files names and identifiers for the MultiQC report to improve display
  • Implement custom sample ordering for MultiQC report
  • Use Ubuntu 20.04 from public ECR as base Docker image

1.2.4 (18 Nov 2021)

1.2.3 - 1.2.4 diff

  • Simplify MultiQC installation (pr75)
    • install MultiQC as conda package
    • install MultiQC_bcbio from umccr org
  • Handle missing PURPLE rainfall plot when >100K variants (iss72)
  • Remove vcf2maf warnings (iss61)
  • Enable MutationalPatterns DBS module (iss59)
  • Render cancer report even with errors (pr69)
  • Adapt unsafe YAML loader (iss63, ext. pr2):
    • bump pyyaml to 5.4.1
    • bump ngs_utils to 2.8.10

1.2.3 (31 Aug 2021)

1.2.2 - 1.2.3 diff

  • Update Dockerfile and install.sh to resolve Snakemake warnings regarding input filepaths containing double slashes '//'
  • Replace root user in Docker container with non-root user

1.2.2 (16 Aug 2021)

1.2.1 - 1.2.2 diff

  • Update NGS_utils to allow custom TSV input (see https://github.com/umccr/NGS_Utils/commit/07e72b4)
  • Pin version of all appropriate Conda packages
  • Explicitly set r-dbplyr version in cancer report Conda env to resolve package conflicts and install issues
  • Use the UMCCR Anaconda channel
  • Replace ECR Ubuntu Docker image base with equivalent from Docker Hub
  • Update (tsvtools)[https://github.com/vladsaveliev/tsvtools] to 0.2.0; Conda package for 0.1.0 no longer exists

1.2.1 (16 Dec 2020)

1.1.3 - 1.2.1 diff

  • Decrease number of predisposition genes to 78 (see vladsavelyev/NGS_Utils#3)
    • Should see a ~60-70% decrease in CPSR variants
  • Disable MutationalPatterns DBS section (see #59)
  • Pierian: revert PURPLE CNV column names to *AllelePloidy (instead of *AlleleCopyNumber) (see #58)
  • Update Hartwig hmftools
    • PURPLE: from 2.43 to 2.51
    • COBALT: from 1.8 to 1.11
    • AMBER: from 3.3 to 3.5
  • DVC: pin to v1.9.1 (see arteria-dev Slack channel 2020-Nov-25)
  • Cancer report updates:
    • update to MutationalPatterns v3
    • use MutationalPatterns gpgr wrappers
    • show MutationalPatterns DBS/INDEL signatures & contributions
    • merge MutationalPatterns Strand Bias plots
    • signatures: exclude artefact SBS sigs (#57)
    • delete sig png files (handled via gpgr)
    • simplify GenomicFeatures::genes call
    • add kataegis table description + more cols
    • add export to excel button for tables
    • fix qc summary table
    • update PURPLE QC metric descriptions
    • bolden h2/h3 report headers
    • add SV BND plot faceted by tier
    • include more conda pkgs in main addendum table
    • navbar expansion
    • export HRD tables nicely
    • add SV 'Map' table description + example
    • SV tables: paste tier/toptier together
    • fix SV column description tables
    • add SV summary description
    • Disable MutationalPatterns DBS section (see #59)

1.1.3 (17 Nov 2020)

1.1.2 - 1.1.3 diff

Mostly moving functionality from the Rmd to gpgr (https://github.com/umccr/gpgr):

  • cancer_report.Rmd:

    • change title to have sample name first (shows up in the browser tab so becomes easier to distinguish)
    • include more inputs as mentioned below
    • include kataegis regions from PURPLE (taken from reannotated PURPLE SNV VCF)
    • include SV BND plots for SR/PR counts
    • split SR/PR column into two for easier sorting for curators
    • use knitr's eval parameter for cleaner chunk evaluation based on presence/absence of SVs
    • move lonely css to css file
    • move AF functionality to gpgr
    • move SV functionality to gpgr
    • move PURPLE functionality to gpgr
    • use gt for summary tables
    • fix yucky HRD tables with gt
    • cnv summary includes germline calls
    • use blank_lines() function for nicer vertical spacing

  • include following R pkgs in the report:

    • details (for writing collapsible description chunks)
    • gt (for tables)
    • jsonlite (for writing tables to json)
    • patchwork (for rearranging plots)

  • file cleanup: delete umccrise .snakemake/metadata directory onsuccess; delete tmp files from rmd report; tar indexcov png + html files. This should save in excess of 200-300 file objects from getting stored on S3.

  • structural.smk: including ALT so that we can grab the mate chromosome when it's been filtered out

  • somatic.smk: include pierian rule, which simply renames SNV/SV/CNV files and subsets SNVs to tumor sample

  • circos_baf.conf/purple.smk: this fixes a long-standing bug (which has already been fixed in PURPLE) where the hg19 karyotype was being used instead of hg38, causing SVs in telomeric regions to get dropped from our custom 'BAF' circos plot.

  • purple.smk: rename variables, bring in germline cnvs, bring in somatic VCF for HRD

  • cancer_report.smk: rename and add variables, most importantly result_outdir which is used for writing the final report tables.

  • Update of vcf2maf.pl conda package recipe

    • this solves the WARNING: Unrecognized biotype "lncRNA". Assigning lowest priority! message which spammed the logs. There will still be spam from vcf2mafpl (Use of uninitialized value in list assignment) due to it not liking something in the input VCF.

  • Replace corrupted HG001_GRCh38_GIAB_highconf_CG-IllFB-IllGATKHC-Ion-10X-SOLID_CHROM1-X_v.3.3.2_highconf_nosomaticdel_noCENorHET7.bed.gz.tbi in reference data

    • this solves an issue where we were getting fewer filtered somatic SNVs due to them not hitting the HMF GIAB BED regions.

  • Pinning gpgr to v0.0.13

1.1.2 (9 Oct 2020)

1.1.1 - 1.1.2 diff

  • Added gpgr dependency to handle R-related enhancements more efficiently
  • Include HRD results from HRDetect and CHORD into the cancer report

1.1.1 (9 Oct 2020)

1.1.0 - 1.1.1 diff

  • Update PCGR to 0.9.0, CPSR to 0.6.0. Includes cosmetic changes and database updates. CPSR counts intronic variants in the "Other variants" summary, as the number there will be high compared to the previous version, however they don't go into any TIER down the report.

1.1.0 (8 Oct 2020)

1.0.9 - 1.1.0 diff

  • Use HMF PoN for annotating/filtering small variants (see vcf_stuff changes).
  • Use DVC for managing the reference data. Add umccrise_refdata_pull script to pull genomes as long as you are signed in under an UMCCR AWS account
  • Support TSV input (if input file ends with .tsv, assumes it's a TSV file with a header and any of the following columns in arbitrary order:
    • sample
    • wgs (WGS tumor BAM, required)
    • normal (WGS normal BAM, required)
    • exome (optional tumor BAM)
    • exome_normal (optional normal BAM)
    • rna (optional WTS BAM, however required for neoantigens)
    • rna_bcbio (optional path to RNAseq bcbio workflow, required for neoantigens)
    • rna_sample (sample name in the RNAseq bcbio workflow above, required for neoantigens)
    • somatic_vcf (tumor/normal somatic VCF calls. If not provided, SAGE will be run)
    • germline_vcf (germline variant calls)
    • sv_vcf (SV calls)
  • Add samtools stats (for TSV runs, where bcbio or Dragen QC is not available)
  • Integrate neoantigens, trigger with -S neoantigens
  • Integrate Optitype HLA calling (used for neoantigen calling, -S neoantigens)
  • Support RNAseq input (currently a part of the TSV input support; used for the neoantigen calling)
  • Oncoviruses: removed alt chromosomes from the reference build used for the integration site calling, which allows for a higher sensitivity in some edge cases

1.0.9 (16 Jul 2020)

1.0.4 - 1.0.9 diff

  • Oncoviruses: decreased the significance threshold to report candidate viruses
  • MultiQC: fix missing variant substitutions plots
  • PCGR: copy the PCGR TSV file into small_variants

1.0.4 (1 Jul 2020)

1.0.0 - 1.0.4 diff

  • Minor fixes:
    • Fix showing conda package versions
    • Oncoviruses section in cancer report: fix showing read support
    • Oncoviruses section in cancer report: rename Pair count -> Read support
    • MultiQC: fix showing the filtered variants column
    • Cancer report: suppress rainfall plot debug prints
  • Add PURPLE summary into MultiQC
  • MultiQC: update the background data
  • Coverage stage works for MacOS without docker (mosdepth available for MacOS)
  • Codebuild: revert back the versioned reference data

1.0.0 (14 Jun 2020):

0.17.12 - 1.0.0 diff

  • Integrated oncoviruses. It detects oncoviral content and possible integration sites, as well as genes affected by integration. Reported viral strains in MultiQC and integration sites and affected genes in the cancer report.
  • More DRAGEN input options:
    • Support tumor and normal sample names different from the output prefix (read from the --RGSM tags)
    • Support multi-fastq runs (when input fastq files and RGSM are specified in CSV files and put in the DRAGEN outout folder by the UMCCR worklow).
  • Custom input. You can also feed custom files using multiple positional arguments. VCF and BAM files are supported. Sample name will be extracted from VCF and BAM headers. For now, VCF is assumed to contain T/N somatic small variant calls, BAM file is assumed to be from tumor. Example:
umccrise umccrise sample1.bam sample2.bam sample1.vcf.gz sample3.vcf.gz -o umccrised -j10
  • Improved parallelization. Re-engineered how the number of CPUs is determined, so when we have many cores available, they are not wasted on a single job that doesn't benefit from many threads. Say, PURPLE COBALT optional number is around 10-15, so if we have 28 cores, we will run COBALT and AMBER in parallel with 14 cores each. This should work for local/AWS runs as well as the HPC runs.
  • PURPLE updated:
    • Use PURPLE v2.40 that reports tumor mutational load and burden
    • Update COBALT GC profile reference file so it fixes a bug ("We have found an issue with the mappability file which is used by COBALT to decide which windows to fit. The file was created on the full set ref genome, when we should have used the analysis set. The impact is that some large (and important) sections of hg38 which have alt contigs are assigned 0 mappability and filtered by COBALT. Particularly the HLA and ERBB2 regions are affected. We have generated new files already and I have checked them, but we are just testing on some real samples.")
  • Integrated OptiType for HLA typing (disabled by default, can be enabled with -T immuno)
  • Re-engineered usage of stages. Add --include-stage (-T) and --exclude-stages (-E), allowed to be provided multiple times. Stages as positional arguments are no longer supported. Instead of:
umccrise /bcbio/final/ multiqc structural

Use:

umccrise /bcbio/final/ -T multiqc -T structural

Positional arguments can now be used for extra custom inputs - BAMs or VCFs, e.g.:

umccrise sample1.bam sample2.bam sample1.vcf.gz sample3.vcf.gz

Sample name will be extracted from VCF and BAM headers. For now, VCF is assumed to contain T/N somatic small variant calls, BAM file is assumed to be from tumor.

In concordance with stages, sample and batch inclusion/exclusion parameters (-s and -e) are now implemented in the same way allowing to be provided multiple times instead of a comma-separated string. Example:

# of all samples in a project, takes only sample1 and sample3, plus all corresponding normal/tumor matches:
umccrise /input/project/final -s sample1 -s sample3

# takes all samples in a project, excluding sample1 and sample2 and corresponding normal/tumor matches:
umccrise /input/project -e sample1 -e sample2

Also add stages: cpsr (decoupled from pcgr), somatic, maf and germline (fine-grained small_variants), mosdepth, goleft and cacao (fine-grained coverage), oncoviruses, microbiome, immuno (novel stages).

Not that some of the stages are inter-dependent, e.g. germline uses somatic to get extra germline leakage, coverage uses purple to get purity-adjusted coverage thresholds, multiqc reports results of conpair, somatic, germline and oncoviral stages, pcgr uses purple to report per-variant purity and ploidy. However, if you target a certain stage specifically and exclude others, umccrise will use the target stage without the dependency. E.g. germline will not contain germline leakage from somatic variants if somatic stage is not included, coverage report will not pull purple and will use reasonable purity threshold defaults instead, multiqc will report on available QC data only, etc.

  • Support reference data location (or prefix) in form of an s3:// or gds:// URL. E.g.
umccrise /input --genomes s3://umccr-refdata-dev/genomes

Would check s3://umccr-refdata-dev/genomes_102, s3://umccr-refdata-dev/genomes_10, and s3://umccr-refdata-dev/genomes, assuming that the reference_data package version is 1.0.2.

Downloads the reference data locally into a ~/umccrise_genomes.

  • Germline variants: updated to CPSR 0.5.2.6, which has an option to pass the predispose coding genes cooridantes instead of a list of genes. Passing the same BED file used to generate -germline.predispose_genes.vcf.gz, which harmonizes this VCF with the CPSR report, and avoids filtering extra variants by CPSR itself. Visible result: extra variants shown in CPSR reports.

0.17.12, 0.17.11 (4 May 2020)

  • Dropping sv-prioritize filters from bcbio, rescuing SVs wrongly marked as "intergenic"
  • CPSR: hiding the genes pane which causes random bugs

0.17.10, 0.17.9 (26 Mar 2020)

  • Fix SV processing for FFPE

0.17.8 (23 Mar 2020)

  • Fix the core usage inside of a Docker image, so COBALT and AMBER will benefit from multiple cores and AWS runs will get much faster now (thanks Peter for spotting the issue)

0.17.7 (14 Mar 2020)

  • CPSR: when prioritizing a mutation effects, always pick the transcripts where the mutations affects the protein.

0.17.6 (12 Mar 2020)

  • To make sure CodeBuild builds code from branches, modify install.sh so it doesn't clone the repo, and move Dockerfile to the root to enable copying the source.

0.17.5 (12 Mar 2020)

  • Fix germline variants count in MultiQC

0.17.4 (11 Mar 2020)

  • Pin PCGR and CPSR and show their correct versions.
  • Pin MultiQC to vladsaveliev/MultiQC rather than umccr-illumina/MultiQC

0.17.3 (9 Mar 2020)

0.17.2 (6 Mar 2020)

Fixes

  • Fix PURPLE memory requirements for single-core runs
  • Remove entrypoint from the Dockerfile

0.17.1 (27 Feb 2020)

  • Fix PCGR and CPSR transcript prioritization issues sigven/cpsr#26
  • Output SAGE results into {batch}/small_variants/sage/{batch}-sage.vcf.gz

0.17.0 (1 Feb 202)

  • Massive cancer report updates by Peter Diakumis:
    • Added new somatic signatures
    • SV tables restrucurised and got much more readable
  • Support DRAGEN project on input instead of bcbio

0.16.0 (31 Dec 2019)

  • Add CodeBuild CI support (thanks @brainstorm for the great work!)
  • Add hg38 MultiQC background samples
  • Update PURPLE, add new plots/stats into the cancer report
  • FFPE pipeline: remove purple-inferred SVs
  • SV prioritization: move filtering BNDs with SR > PR from Rmd into the prioritization script (so this filter reflects in VCF)
  • Added GRIDSS/PURPLE/LINX wrapper (gpl) and env (umccrise_hmf)
  • hpc_utils: refactoring of genomes_dir detection functions. Add set_genomes_dir
  • Fix multithreading when running in a Docker
  • Support Gadi
  • Update PCGR to 0.8.4 and CPSR to 0.5.2:
  • Support unprioritized SV VCFs naming (when sv-prioritize is disabled in bcbio)

0.15.6 (28 Aug 2019)

  • SV prioritization:
    • prioritize CN=0 events in tumor suppressors genes
    • fix transcript order in PCGR (use Rank first)
    • report all (principal+alt) transcripts comma-separated
    • skip interaction events
  • Cancer report SV table:
    • Add SV event number
    • Fix ref/alt read support labels
  • Cancer report: reorganize blocks, increase report width
  • SV VCF: split samples, keep tumor (for Pierian)
  • Add MultiQC Docker container for bcl2fastq and interop

0.15 (21 May 2019)

  • Update PCGR to 0.8.1 and CPSR to 0.4.1 (with a new scheme for germline filtering)
  • Pass tumor purity and ploidy into PCGR
  • VEP annotation pick order: pick APPRIS first instead of canonical (to match with SV)
  • Add the detailed workflow description

0.14 (1 Apr 2019)

  • Update SV prioritization method
  • New SV table in the rmd report
  • New PURPLE tools, PURPLE annotates and rescues SVs
  • Panel of normals rebuilt with lowfreq tumor-only Mutect2, threshold is raised to 5
  • Coverage report threshold: adjust to purity (CACAO and mosdepth)
  • Support of Cromwell MultiQC
  • Report somatic germine leakage (gnomAD variants) in CPSR

0.13 (15 Feb 2019)

  • Re-enable CACAO
  • Support Cromwell runs
  • MultiQC:
    • add somatic and germline filtering stats
    • run bcftools-stats for filtered calls
    • sort samples in tumor-normal order
    • add columns: secondary alignments, MQ=0 reads, viral
    • clean up
  • Integrate SAGE: to refine variant calls in hotspot locations
  • SVs:
  • FFPE:
    • Hard subset small variants to non-gnomad if >500k, subset to cancer genes if doesn't help
    • Hard subset SVs to non-intergenic if >1000, subset to prioritized if doesn't help
    • Rerun BPI on priority SVs when disabled in bcbio
  • Cancer report significantly updated (thanks to Peter's work!):
    • Add mutational signatures descriptions
    • Discontinue bookdown for a speed boost
    • Restructure tables and plots
    • Add explanations how files are generated in the pipeline

0.12 (23 Jan 2019)

  • Somatic annotation/filtering:
    • Filter VarDict strand bias and HP indels
    • Abstract somatic annotation into somatic_vcf_annotate script
    • Add more sources: HMF mappability, ICGC, COSMIC, TCGA hotspots, ENCODE, LCR
  • Populate VD and AD in final VCFs (update to pcgr_prep)
  • Support Strelka2-only runs, populate Strelka2 AD for Purple
  • Correct threading for cluster runs
  • Key genes:
    • Update umccr cancer genes to 2019 (remove TSG, add cancermine and NGC)
    • Use NGC+cancermine oncogene/tsg defenitions
    • Add predispose BED files for germline runs
  • Update Purple to sambamba-free Amber
  • Integrate Conpair (replaces VerifyBAMID and Peddy)

0.11 (6 Dec 2018)

  • Add Per-sample MultiQC reports with reference samples
  • Add Sig.rmd Purple table
  • Add summary Purple-vs-bcbio comparison

0.10 (30 Nov 2018)

  • SNV filtering improvements:
    • Assuming calling with 1%
    • First round of PoN filter: remove with >=3 hits
    • Annotate with PCGR and keep all tier 1-3
    • Tier 4 and noncoding - filter with:
      • gnomad_AF <=0.02
      • PoN hits >=1
      • indels in bad promoters
      • DP<25 & AF<5% in tricky regions or GiaB LCR
  • Integrate CACAO (temporary disabled)

0.9 (15 Nov 2018)

  • Integrate Purple
  • Integrate CPSR (PCGR reports for germline variants)
  • Update PCGR
  • Use Arthur's new list of cancer key genes for minibams and coverage
  • Use Sigve's+Sean's predispodition gene lists for germline reporting

0.8 (7 Aug 2018)

  • SV filtering: require all of FILTER fields to be PASS or rejected by sv-prioritizer
  • SV filtering: always keep SR or PR above 10x regardless of BPI_AF (but above 5x only if BPI_AF>10%)
  • SV filtering: check BPI_AF in both breakends
  • Add option --no-s3 to skip IGV upload; use it in tests
  • Restart failed jobs tree times (helps with PCGR random issues)
  • Exlcuding/including samples: perform before check for correct YAML
  • Exlcuding/including samples: support {batch}__{sample} identifiers
  • Sopport the new and old germline VCF bcbio naming (-.vcf.gz and -germline-.vcf.gz)

0.7 (26 Jun 2018)

  • Dockerized version with --docker
  • Custom reference data with --pon, --ref-fasta, --truth-regions, or --bcbio-genomes

0.6 (15 May 2018)

  • Use snakemake groups to optimize execution on cluster
  • Use submission wrapper to automate cluster resources
  • Filter structural variants: BPI_AF>=10% and read support>=5
  • Keep sv-prioritize hard-filtered variants
  • 0.6.1

0.5 (7 May 2018)

  • PCGR is deployed directly on Spartan, so no AWS dependency.
  • Add PCGR wrapper: pcgr variants.vcf.gz cnv.tsv -o results [-g hg38]
  • Correctly providing memory resources on HPC to avoid oom-kill
  • On Spartan, support --cluster-auto to automatically substitute proper cluster parameters

0.4 (17 Apr 2018)

  • Propagate snakemake's cluster options to the wrapper
  • Propagate snakemake's --rerun-incomplete and to the wrapper
  • PCGR: stop using the UID, skip uploading if already uploaded
  • Fix HTTP_PROXY setting for running on Spartan worker nodes
  • New panel of normals: add the large A5 cohort, normalize normals, check only POS for indels
  • Correctly set threads to mibibams samtools runs

0.3 (17 Jan 2018)

  • Refactor output folder structure (see docs)
  • pcgr_download taget to automatically pull results
  • Automatically attempt downloading PCGR results after minibams
  • Minibams always run in the ned
  • Use --unlock to automate the snakemake unlock pattern to continue unterrupted runs
  • Move vcf_stuff to a separate repo
  • Remove GL* chromosomes from the cnvkit diagram
  • PCGR tomls: changed mutsignatures_normalization to genome
  • Rmd: add SV-prioritize table