Version history

Added

  1. Added support for multiple FASTQ files for the same sample. These will be merged at the start of the pipeline #821
  2. Added a stub test for merging single-end FASTQ files of the same sample #848
  3. Added --skip_gencode_filter to build the CTAT genome library without STAR-Fusion’s --human_gencode_filter, for annotations that are not GENCODE #832
  4. GATK4_MARKDUPLICATES now creates and publishes an index for the duplicate-marked BAM #784

Changed

  1. Moved the pipeline to full strict syntax compatibility
  2. Updated the minimal Nextflow version to 26.04.0
  3. Bumped to nf-core template v4.2.2
  4. Moved references from s3://nf-core-awsmegatests/rnafusion/references to s3://ngi-igenomes/nf-core/pipelines/rnafusion/references
  5. Added --no-fusiongdb2 to the fusionreport download command since the fusiongdb2 endpoint no longer exists. This will be removed once fusionreport is updated to address this issue.
  6. The samplesheet no longer requires unique sample values: rows with the same sample are combined into one sample #821
  7. Bumped nf-schema to 2.7.3
  8. Reorganised the parameter schema so inputs, references and alignment options are grouped together. This only changes how parameters are presented. Based on the original reorganisation by FriederikeHanssen in #782
  9. Added --no-fusiongdb2 to the fusionreport download command since the fusiongdb2 endpoint no longer exists. FUSIONREPORT_DOWNLOAD (patched) now creates an empty fusiongdb2 database instead, because fusion_report run always queries it and failed with no such table: fusiongdb2 on references built with this release #842. This will be removed once fusionreport is updated to address this issue.
  10. Updated the nf-core modules. SALMON_INDEX and SALMON_QUANT stay on 1.10.3 because salmon 2.x rejects the published index, and utils_nfschema_plugin stays on the nf-schema 2 version
  11. Replaced the local GENCODE_DOWNLOAD and HGNC_DOWNLOAD modules with the nf-core wget and gunzip modules #575.
  12. Cleaned up pipeline-level lint findings: removed the deprecated when: section from the local VCF_COLLECT module, added meta.yml files to the local subworkflows, filled in the pipeline-specific contribution guidelines and ignored the renamed iGenomes config files in the files_exist lint test
  13. Updated the starfusion/detect module to include its new nf-test with a chimeric junction file as input #593

Fixed

  1. Fixed the PL tag of the STAR read group being merged into the SM tag when a sequencing platform is set #822
  2. Fixed single-end FASTQ samplesheets being rejected by the input schema since v4.0.0 #860
  3. Fixed --references_only requiring --input, which broke the documented reference-building command #861
  4. Fixed the test_build profile failing on the STAR-Fusion tutorial minigenome by setting --skip_gencode_filter, since that annotation is not GENCODE #832
  5. Fixed samples spread over several samplesheet rows keeping only the first row’s meta: seq_center/seq_platform from other rows are kept (and must agree), and rows without FASTQ files no longer break the single-end/paired-end check #847
  6. Fixed every run checking the remote dfam, pfam, AnnotFilterRule, cancer introns and fusion annotation library defaults at launch. They are now only resolved when the STAR-Fusion reference is built #843
  7. Fixed the completion email arguments being shifted by one, and the MultiQC report being passed to the email as a [meta, report] tuple, which stopped the report from being attached #849
  8. Fixed the pfam warning being printed on every run, because --pfam_file was compared as a path instead of a URL #845
  9. Fixed MultiQC regressions: a custom --multiqc_config is applied on top of the pipeline config again instead of replacing it, MultiQC is skipped again with --skip_qc and --references_only, and the MultiQC meta id is rnafusion #844
  10. Fixed strandedness from the samplesheet not being added to the meta map, so PICARD_COLLECTRNASEQMETRICS never got the right --STRAND_SPECIFICITY and the per-sample strandedness check never ran #856
  11. Fixed ctatsplicing failing on a prebuilt STAR-Fusion reference without a cancer splicing index. The index is now added to a copy of the reference, which is published to references/starfusion in --outdir #704
  12. Fixed additional STAR parameters not being correctly passed to FusionInspector by updating the module #841
  13. Fixed ARRIBA_VISUALISATION crashing on a samplesheet BAM without an index. The index is now created when ctatsplicing or fusioninspector need it, and a sample without confirmed fusions now logs a warning that its visualisation is skipped #780
  14. Fixed prebuilt references being rebuilt because a directory was judged empty from a single file, e.g. an empty checkpoint file in the STAR-Fusion __chkpts/ folder. A reference directory now counts as present if any file in it has content, and an empty directory no longer crashes the check #774
  15. Increased the memory of FUSIONCATCHER_FUSIONCATCHER to 48 GB, since bbduk.sh could run out of memory with the process_medium default #773

Tool version updates

Tool Old version New version
agat 1.5.1 1.6.1
ctatsplicing 0.0.2 0.0.3
fastp 1.1.0 1.4.0
fastqc 0.12.1 0.13.0
fusion-report 4.1.1 4.1.2
gatk4 4.6.1.0 4.7.0.0
multiqc 1.33 1.35
picard 3.3.0 3.5.0
samtools 1.21, 1.22.1, 1.23.1 1.25
stringtie 2.2.1, 2.2.3 3.0.3
ucsc 447 482

What’s Changed

New Contributors

Full Changelog: https://github.com/nf-core/rnafusion/compare/4.1.3...4.2.0

What’s Changed

New Contributors

Full Changelog: https://github.com/nf-core/rnafusion/compare/4.1.0...4.1.1

Added

  • Added a BAM/CRAM conversion to FASTQ step to run tools that depend on FASTQ files. It still is advised to supply a FASTQ file if possible to reduce runtime of the pipeline. #783

Fixed

  • Fix issue with vcf_collect when no fusions are found by fusioninspector #786

Patch fix for vcf_collect.py in cases when FusionInspector is empty but staying compatible with references from 3.0.1

Major changes include

  • Recommended download of reference with aws --no-sign-request s3 sync s3://nf-core-awsmegatests/rnafusion/references/ <path_to_references>
  • Changed reference fasta file to the primary assembly, moved from ensembl 102 to gencode 46
  • New option to avoid using COSMIC --no_cosmic
  • New option to skip generating the VCF summary: --skip_vcf
  • New tool: CTAT-Splicing to detect cancer splicing events
  • Introduced support to start from CRAM/BAM files and (splice) junctions files
  • Fusion-report score is now called the fusion indication index (FII)
  • Extra trimming for FusionCatcher in case reads are larger than 100 bp
  • Single alignment for arriba and STAR-Fusion
  • Use of a parameter --tools STRING to select which tools to enable

Bug fixes.

Changelog:

Changed

Fixed

Patch to fix software versions output

Main changes:

  • Squid and pizzly are no longuer supported
  • Removed --fusioninspector_filter and --fusionreport_filter in favor of --tools_cutoff (default = 1, no filters applied)
  • Removed FusionGDB from fusion-report
  • Improved vcf file creation, using --annotate --examine_coding_effect with fusioninspector to collect more data
  • Removed harsh trimming option, fastp trimming still available
  • Removed qualimap rna_seq

Bug fixes:

  • building references is possible without a samplesheet
  • Fix channel i/o issue in StringTie workflow
  • Various fusion-report issues with download
  • Improved CI

Changelog:

Changed

  • Use institutional configs by default #381
  • Remove redundant indexing in starfusion and qc workflows #387
  • Output bai files in same directory as bam files #387
  • Update and review documentation #396
  • Update picard container for PICARD_COLLECTRNASEQMETRICS to 3.0.0 #395
  • Renamed output files #395
    • Arriba visualisation pdf from meta.id to meta.id_combined_fusions_arriba_visualisation
    • cram file from output bam of STAR_FOR_ARRIBA: meta.id to meta.id_star_for_arriba
    • cram file from output bam of STAR_FOR_STARFUSION: meta.id to meta.id.star_for_starfusion.Aligned.sortedByCoord.out
    • fusion-report index.html file to meta.id_fusionreport_index.html
    • meta.id.vcf output from MEGAFUSION to meta.id_fusion_data.vcf

Fixed

  • Tail trimming for reverse reads #379
  • Set html files as optional in fusionreport #380
  • Provide gene count file by default when running STAR_FOR_STARFUSION #385
  • Fix fusion-report issue with MACOXS directories #386
  • The fusion lists is updated to contain two branches, one in case no fusions are detected and one for if fusions are detected, that will be used to feed to fusioninspector, megafusion, arriba visualisation #388
  • Update fusionreport to 2.1.5p4 to fix 403 error in downloading databases #403

Removed

  • samtools sort and samtools index for arriba workflow were dispensable and were removed #395
  • Removed trimmed fastqc report from multiqc #394

New Contributors

Full Changelog: https://github.com/nf-core/rnafusion/compare/2.3.4...2.4.0

Changed permissions on megafusion script

Added

  • Shell specification to bash
  • COSMIC password put into quotes
  • Trimmed reads QC in MultiQC
  • Add ARRIBA_VISUALISATION to processed affected by --skip_vis
  • Option fusionreport_filter to in/activate fusionreport displaying of fusions detected by 2 or more tools

Changed

  • Arriba visualisation now runs for FusionInspector (combined tools) results, not only Arriba results
  • Updated metro map with trimming options and placed Arriba visualisation after FusionInspector
  • Exit with error when using squid in combination with any ensembl version different from 102

Fixed

  • Channel issue with indexing of files with using --cram squid
  • Arriba references published in the correct folder

Version 2.0.0 includes stringtie in --all, a fastp trimming option, fixed for arriba references among other (see details below).

Important: the references need to be rebuilt.

Check usage.md for how to use the new options.

Detailed CHANGELOG:

Added

  • exitStatus 140 now part of the retry strategy
  • stubs to all local modules
  • --stringtie option added with StringTie v2.2.1 to detect splicing events. Not included in fusion-report or fusionInspector summaries. Included in the --all workflow
  • Generation of ribosomal RNA interval list with build_references and use it in picard CollectRnaMetrics
  • Add csv output to fusionreport
  • Trimming workflow using fastp: use trimmed reads for all tools
  • whitelist parameter to add custom fusions to the detected ones and consider the whole for the fusionInspector analysis
  • Compression to CRAM files for arriba, squid and starfusion workflows (fusioncatcher and pizzly do not produce SAM/BAM files, fusioninspector BAM files are too small to benefit from compression)
  • --qiagen option to download from QIAGEN instead of COSMIC (use QIAGEN user and password for cosmic_username and cosmic_passwd)
  • Bumped STAR genomegenerate time request for building as it was always crashing for most users
  • Fixed issue with arriba visualisation parameters #326

Changed

  • Test profiles unified under ‘test’ but if the references do not all need to be downloaded, run with -stub
  • Update CUSTOM_DUMPSOFTWAREVERSIONS to use multiqc version 1.13
  • Updated to nf-core template 2.7.2, with all module updates
  • MultiQC updated to 1.13a in process dumpsoftwareversion
  • Patch fusion-report version with fixed mittelman DB and DB extraction date written into software_version.yaml
  • Arriba references back to downloading with build_references instead of taking from container
  • Arriba visualisation now running with Arriba v2.3.0
  • Updated STAR-Fusion to 1.12.0

Fixed

  • AWS megatest to display on nf-core website
  • arriba visualisation references updated to 2.3.0
  • Removed issue with multiple outputs in samtools view for squid

Removed

  • FUSIONINSPECTOR_DEV process as the option fusioninspector_limitSjdbInsertNsj is part of the main starfusion release

IMPORTANT Since STAR has been updated, a new STAR index needs to be built as reference. nextflow run nf-core/rnafusion --genomes_base /PATH/TO/REFERENCES --starindex --build_references --outdir OUTDIR -profile singularity or docker

Added

  • FusionCatcher single_end support for single reads ABOVE 130 bp
  • --fusioninspector_only parameter to run FusionInspector standalone feeding gene list manually with parameter --fusioninspector_fusions PATH
  • --fusioncatcher_limitSjdbInsertNsj parameter to feed –limitSjdbInsertNsj to FusionCatcher
  • --fusioninspector_limitSjdbInsertNsj parameter to feed –limitSjdbInsertNsj to FusionInspector !!Any other value than default will use the dev version of FusionInspector!!
  • OPTIONAL trimming option --trim for hard trimming to 75 bp in case of high read-through. Only fusioncatcher uses trimmed reads as STAR-based fusion detection tools are less sensitive to read-through
  • picard metrics, STAR final log, and QualiMap output included in MultiQC report

Changed

  • seq_platform and seq_center changed from boolean to string
  • seq_platform set to an empty string and seq_center set to an empty string if not existing
  • Arriba use ensembl references-built starindex independently of starfusion_build parameter
  • ftp to http protocol for STARFUSION_BUILD process Pfam-A.hmm.gz download as ftp causes issues on some servers
  • Updated README and usage documentation with more detailed information and metro map
  • Arriba use ensembl references-built starindex independently of starfusion_build parameter
  • Update of the single-end reads support table in README, added recommendation to use single-end reads only in last resort
  • STAR updated to 2.7.10a
  • Arriba updated to 2.3.0, references for blacklist and protein domains changed to 2.3.0 from singularity/docker container -> arriba download of references not necessary any more
  • multiQC updated to 1.13a
  • picard updated to 2.27.4
  • dumpsoftwareversions module updated to use multiqc=1.12 containers

Fixed

  • FusionInspector does not mix sample reads with fusion lists and meta information from other samples anymore
  • Arriba visualisation does not mix sample reads with fusion lists and meta information from other samples anymore
  • logging of STAR-fusion and fusionreport version

The usage of the pipeline has been updated. It is recommended to reads the usage.md document.

Added

  • Added qualimap/rnaseq v2.2.2d from nf-core modules
  • Added UCSC gtfToGenePred v377
  • Added picard CollectRnaSeqMetrics v2.26.10
  • Added picard MarkDuplicates v2.26.10 from nf-core modules
  • Added cat/fastqc from nf-core modules
  • Added possibility for manually feeding the results of fusions from different tools to speed-up reruns
  • STAR-Fusion references can be downloaded or built but downloaded references are NOT RECOMMENDED as not thoroughly tested (–starfusion_build parameter is true by default, use –starfusion_build false to use downloaded STAR-Fusion references).

Changed

  • Upgrade default ensembl version to 102
  • Upgrade to nf-core/tools v2.3.2
  • Upgrade Arriba v1.2.0 to Arriba v2.2.1
  • Upgrade FusionCatcher v1.20 to FusionCatcher v1.33
  • Upgrade STAR-fusion v1.8.1 to STAR-fusion v1.10.1
  • Upgrade STAR v2.7.1 to STAR v2.7.9
  • Upgrade fusion-report v2.1.3 to fusion-report v2.1.5
  • Upgrade kallisto v0.44.0 to kallisto v0.46.2
  • Upgrade fastqc v0.11.8 to fastqc v0.11.9
  • Upgrade samtools v1.9 to samtools v1.15.1
  • Upgrade arriba references from v1.2.0 to v2.1.0
  • Upgrade fusioncatcher references from v98 to v102
  • Use arriba (detect only), kallisto and STAR from nf-core modules
  • Instead of separate script to build the references, added –build_references argument in the main –fasta argument is not required with –build_references and set by default to the ensembl references built in the detection workflow

Removed

  • Ericscript tool
  • GRCh37 support. Subdirectory with params.genome are removed
  • Running with conda

[1.2.0] nfcore/rnafusion - 2020/07/15

  • Fusion gene detection tools:
    • Arriba v1.2.0
    • Ericscript v0.5.5
    • Fusioncatcher v1.20
    • Pizzly v0.37.3
    • Squid v1.5
    • STAR-Fusion v1.8.1
  • Visualization tools:
    • Arriba v1.2.0
    • FusionInspector v2.2.1
  • Other tools:
    • fusion-report v2.1.3
    • FastQ v0.11.8
    • MultiQC v1.7
    • STAR aligner v2.7.1a

Added

  • Added social preview image #107
  • Simplified running the pipeline, better documentation

Changed

  • Upgrade fusion-report v2.1.2 to fusion-report v2.1.3
  • Upgrade fusion-report v2.1.1 to fusion-report v2.1.2
  • Upgrade fusion-report v2.1.0 to fusion-report v2.1.1
  • Upgrade Arriba v1.1.0 to Arriba v1.2.0
  • Upgrade fusion-report v2.0.2 to fusion-report v2.1.0

Fixed

  • Missing strip-components in download-references.nf/star-fusion #148
  • Missing version prefix for cdna #143
  • samtools missing header in empty file for FusionInspector ref
  • Removed profile from helper scripts #139
  • Wrong url path for Pfam-A.hmm.gz #140

Removed

  • Removed scripts/download-singularity-img.sh and download-singularity-img.nf as they are not necessary any more

v1.1.0 - 2020/02/10

  • Fusion gene detection tools:
    • Arriba v1.1.0
    • Ericscript v0.5.5
    • Fusioncatcher v1.20
    • Pizzly v0.37.3
    • Squid v1.5
    • STAR-Fusion v1.6.0
  • Visualization tools:
    • Arriba v1.1.0
    • FusionInspector v1.3.1
  • Other tools:
    • fusion-report v2.0.1
    • FastQ v0.11.8
    • MultiQC v1.7
    • STAR aligner v2.7.0f

Added

  • Added Arriba 1.1.0 #63
  • Added Batch mode #54

Changed

  • Updated examples and configurations
  • Upgraded fusion-report v1.0.0 to fusion-report v2.0.1
  • Divided running_tools into fusion and visualization tools
  • Updated STAR in Squid, Fusion-Inspector version to 2.7.0f
  • Upgraded STAR-Fusion v1.5.0 to STAR-Fusion v1.6.0 #83
  • Parameter igenomesIgnore renamed to igenome #81
  • Finished STAR-Fusion file renaming #18
  • Updated logos
  • Updated to nf-core 1.8 TEMPLATE

Fixed

  • iGenomes optional, but not really #91
  • Updated fusioncatcher to latest 1.20 version also solving #95

Removed

  • Variables pizzly_fasta and pizzly_gtf have been removed and replaced with transcript and gtf
  • Jenkisfile, test configuration, pylintrc configuration
  • Removed igenomes.config because the pipeline only supports Ensembl version

v1.0.2 - 2018/05/16

Changed

  • Bumped nf-core template to 1.6 #69

Fixed

  • Fixed COSMIC parameters not wrapped in quotes #75
  • Implemented output output for fusion tools #72
  • Fixed reference download link for STAR-Fusion #71

v1.0.1 - 2018/04/06

Added

  • Added support for extra parameters for tools STAR-Fusion, FusionCatcher and fusion-report
  • Added example configuration for singularity and docker
  • Added fusion-report into the stack #62, #55, #53, #51
  • Added nextflow helper script download-singularity-img.nf
  • Added nextflow helper script download-references.nf
  • Added Jenkinsfile for in-house testing

Changed

  • Updated installation of FusionCatcher (available now on bioconda)

Fixed

  • Fixed empty symlinks (input.X) in fusion-report #68
  • Fixed FASTA issues #60
  • Fixed centralized nf-core/config #64
  • Fixed scrape_software_versions.py to parse tools versions correctly #65

Removed

  • Removed Singularity

Initial release of nf-core/rnafusion 🎉

This is the first production release under nf-core flag previously under SciLifeLab/NGI-RNAfusion.

This version includes:

  • Fusion gene detection tools:
    • STAR-Fusion v1.5.0
    • Fusioncatcher v1.00
    • Ericscript v0.5.5
    • Pizzly v0.37.3
    • Squid v1.5
  • Visualization tools:
    • FusionInspector v1.3.1
  • Other tools:
    • Summary report v1.0
    • FastQ v0.11.8
    • MultiQC v1.7
    • FusionGDB updated at 2019/01/23

Thanks to everyone that contributed to this release 💪 Special thanks go to @MaxUlysse, @alneberg, @ewels, @szilvajuhos and @apeltzer