Dev -> master for nf-core/rnaseq 3.23.0 by pinin4fjords · Pull Request #1717 · nf-core/rnaseq

pinin4fjords · 2026-02-26T09:11:03Z

Release 3.23.0

330 commits, 546 files changed since 3.22.2.

This is a major feature release with new alignment/quantification pathways, prokaryotic support, improved preprocessing options, and many quality-of-life fixes.

Review guide

This is a large release. Rather than reviewing all 330 commits, the component PRs below are grouped by theme. Each was individually reviewed before merging to dev.

New features

PR	Summary	Key files
#1685	Parabricks GPU-accelerated STAR (`--use_parabricks_star`)	`workflows/rnaseq/main.nf`, `conf/modules/parabricks_star.config`
#1688	Prokaryotic RNA-seq support (`-profile prokaryotic`): Bowtie2+Salmon, GFFREAD transcript extraction, STAR CDS config	`workflows/rnaseq/main.nf`, `conf/prokaryotic.config`, `subworkflows/local/prepare_genome/`
#1616	Sylph contamination detection	`workflows/rnaseq/main.nf`, `nextflow_schema.json`
#1664	Multiple rRNA removal tools (`--ribo_removal_tool`): SortMeRNA, Bowtie2, RiboDetector; BBSplit MultiQC reporting	`subworkflows/nf-core/fastq_qc_trim_filter_setstrandedness/`, `workflows/rnaseq/main.nf`
#1697	tximport for RSEM outputs: length-scaled counts, SummarizedExperiment, DESeq2 QC	`workflows/rnaseq/main.nf`, `subworkflows/local/quantify_rsem/`
#1700	Post-filtering FastQC step after BBSplit/rRNA filtering (#1276)	`workflows/rnaseq/main.nf`
#1713	Per-sample seq_platform/seq_center samplesheet columns for BAM read groups	`workflows/rnaseq/nextflow.config`, `assets/schema_input.json`
#1718	`--seq_platform` global parameter for default `PL` read group tag (per-sample samplesheet values take priority)	`workflows/rnaseq/nextflow.config`

Bug fixes and improvements

PR	Summary
#1686	Support uncompressed FASTQ input files (#1343)
#1696	Fix offline mode (`NXF_OFFLINE=true`) by skipping `igenomes_base` validation (#1690)
#1704	Include single-library samples in merged fastq output with `--save_merged_fastq` (#748)
#1707	Enable UMI deduplication with `--aligner star_rsem` (#829)
#1708	Allow `--skip_alignment` with pre-built pseudo-aligner index without `--fasta` (#1706)
#1711	Scope GPU container flags to `process_gpu` only, fixing CPU-only node failures (#1710)
#1716	Informative error when Salmon strandedness inference fails for `auto` samples
#1738	Fix iGenomes STAR conda: use `bioconda` channel for x86_64, reserve `seqera` channel for ARM only

Refactoring and maintenance

PR	Summary
#1665	Bulk module/subworkflow updates; replace CUSTOM_GETCHROMSIZES with SAMTOOLS_FAIDX; ARM containers
#1667	Enhance RSEM output exports
#1669	Enable SeqKit stats MultiQC module for RiboDetector
#1677	Nextflow 25 strict syntax fixes
#1687	Fix PREPARE_GENOME tests, file-names-only snapshots
#1689	Migrate to topic-based version reporting
#1694	Replace local iGenomes STAR modules with nf-core aliases
#1724	Update umitools modules: remove patches, add conda channel prefixes, update Wave containers
#1725	Refine .nftignore patterns to reinstate tx2gene MD5 checking, remove redundant UNTAR version collection
#1735	Code quality improvements: named destructuring, rename `ch_dummy_file`, align GPU CI workflow

Documentation

PR	Summary
#1672	Document star_rsem STAR aligner settings
#1709	Update metro map with nf-metro v0.4.7
#1714	Re-render metro map with nf-metro v0.5.2
#1719	Docs and changelog cleanup: move descriptions above `<details>` blocks, add missing deps
#1728	Re-render metro map with nf-metro v0.5.4

CI fixes

PR	Summary
#1730	Bump Nextflow to 25.04.3 to fix conda `--mkdir` bug (nextflow-io/nextflow#5947)
#1734	Compose `test_prokaryotic` profile on `prokaryotic` to remove duplicated settings
#1736	Fix prokaryotic nf-tests to use the prokaryotic profile instead of duplicating params

New parameters

Parameter	Description
`--ribo_removal_tool`	Choice of rRNA removal tool (sortmerna, bowtie2, ribodetector)
`--sylph_db`	Sylph database for contamination detection
`--sylph_taxonomy`	Sylph taxonomy metadata
`--gffread_transcript_fasta`	Use GFFREAD to extract transcript FASTA from genome
`--extra_bowtie2_align_args`	Extra arguments for Bowtie2 alignment
`--use_parabricks_star`	Use NVIDIA Parabricks for GPU-accelerated STAR alignment
`--seq_platform`	Default sequencing platform for BAM read group `PL` tag

New profiles

Profile	Description
`prokaryotic`	Settings optimized for bacterial/archaeal RNA-seq data

Test plan

CI nf-test passes
Linting passes

🤖 Generated with Claude Code

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

The .ifEmpty(null) caused null values to flow through to processVersionsFromYAML when the versions channel was empty, causing 'Cannot invoke Path.getFileSystem() because path is null' error. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

- Remove workflow.out.versions from test assertions in align_star and quantify_rsem subworkflows (versions now emitted via topics) - Update snapshots to remove versions.yml entries and null placeholders - Fix sentieon_default fq version (0.12.0) and picard version (3.4.0) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

…s entries - Remove trailing null entries from align_star sentieon snapshots - Add BEDTOOLS_GENOMECOV_REV, FQ_LINT_AFTER_BBSPLIT, FQ_LINT_AFTER_TRIMMING to sentieon_default snapshot Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

- Simplify Bowtie2 description in docs/output.md - Add note clarifying transcript/gene-level output for prokaryotes - Add --skip_qualimap and --skip_bigwig to manual config table - Update profile description to include "or analyses" - Add --alignIntronMax 1 for STAR prokaryotic alignment - Enable params.save_unaligned for Bowtie2 (was hardcoded false) - Remove --no-mixed flag from Bowtie2 (valid for operon reads) - Change 'transcriptome' to 'transcripts' for consistency - Change "recommended" to "alternative" for star_salmon option - Add credits and --extra_bowtie2_align_args to CHANGELOG Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Remove .out.versions references for modules that were updated to topic-based version reporting: - KRAKEN2 - BRACKEN - SYLPH_PROFILE - SYLPHTAX_TAXPROF These modules now emit versions via channel.topic('versions') and no longer have the .out.versions channel. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

feat: Migrate to topic-based version reporting

Resolves merge conflicts, keeping prokaryotic support refactoring: - Maintains single BAM_DEDUP_UMI import (vs aliased versions on dev) - Keeps ALIGN_BOWTIE2 for prokaryotic support - Accepts dev's module/subworkflow updates and topic-based version reporting Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

After merging origin/dev with topic-based version reporting, remove incompatible .out.versions calls from: - workflows/rnaseq/main.nf (ALIGN_STAR, FASTQ_ALIGN_HISAT2, BAM_DEDUP_UMI, QUANTIFY_RSEM, ALIGN_BOWTIE2) - subworkflows/local/align_bowtie2/main.nf (remove versions emit) - subworkflows/local/prepare_genome/main.nf (BOWTIE2_BUILD.out.versions) All underlying modules now use topic-based version reporting. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Add --rg-id flag to Bowtie2 alignment to include read group information in BAM output. This is required by Picard MarkDuplicates (3.4.0) which validates read groups before processing. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Update snapshot after topic-based version reporting migration. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

feat: Add prokaryotic RNA-seq support with -profile prokaryotic

…g ignore approach Instead of adding `tag "cpu"` to every existing test file (20+ files changed), use SKIP_PARABRICKS env var in nf-test.config to conditionally ignore GPU tests on CPU runners - consistent with existing SKIP_SENTIEON pattern. - Reset all non-GPU test files to dev state (remove tag "cpu" additions) - Extract parabricks tests from prepare_genome/main.nf.test to separate file - Add SKIP_PARABRICKS env var to nf-test.config ignore logic - Add SKIP_PARABRICKS=true to CPU CI workflow - Fix nf-core/methylseq typo in GPU workflow Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

# Conflicts: # CHANGELOG.md # subworkflows/local/prepare_genome/main.nf

Consolidate ext.args for STAR and Parabricks rnafq2bam into a unified closure in align_star/nextflow.config, with isPbrun branching, read group injection, and zcat flag. Fix process name prefix in prepare_genome config and remove debug echo from rnafq2bam module. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Rewrite pipeline-level and prepare_genome subworkflow tests for the Parabricks path using the standard eukaryotic test dataset instead of the full-size data. Add SKIP_PARABRICKS env var to tests/nextflow.config, update align_star test config and snapshots from GPU CI. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Document the --use_parabricks_star flag in the usage guide, covering requirements, expected behaviour, and known limitations versus CPU STAR. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

The Nextflow linter does not allow `def` variable declarations mixed with config statements in a config block. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

This reverts commit 415bc94.

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

ci: bump Nextflow to 25.04.3 to fix conda --mkdir bug

@adamrtalbot

Make test_prokaryotic include prokaryotic.config first, then layer test-specific settings on top. This removes duplicated prokaryotic params from test_prokaryotic.config (prokaryotic flag, aligner, gffread, featurecounts_feature_type, skip_bigwig, skip_rseqc, skip_dupradar, skip_qualimap) since they are now inherited. The nf-test params blocks in tests/prokaryotic.nf.test are kept as-is since CI runs those tests with -profile test,docker (not test_prokaryotic), so the params block is the mechanism that sets prokaryotic behavior in CI. Addresses review feedback from @adamrtalbot in PR #1717. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

GPU runners are expensive - just test minimum version. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

refactor: code quality improvements from PR #1717 review

…bility refactor: compose test_prokaryotic profile on prokaryotic

@adamrtalbot

Add config "../conf/prokaryotic.config" to the prokaryotic test suite so that prokaryotic-specific settings (prokaryotic flag, aligner, gffread, featurecounts_feature_type, skip_bigwig, skip_rseqc, skip_dupradar, skip_qualimap) come from the profile config rather than being duplicated in each test's params block. Tests now only set: - Test data paths (input, fasta, gff, gtf=null, etc.) - Aligner override for star_salmon tests - Test-specific skips (bbsplit, preseq, biotype_qc, deseq2_qc) - min_trimmed_reads = 0 Addresses review feedback from @adamrtalbot in PR #1717. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

The prokaryotic config now applies skip_bigwig = true to all prokaryotic tests (including star_salmon), which is correct behavior but changes the expected output. Removing snapshots so CI regenerates them. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Remove bigwig-related entries (BEDTOOLS_GENOMECOV_FW/REV, UCSC_BEDCLIP/BEDGRAPHTOBIGWIG, bigwig file paths) from the star_salmon prokaryotic snapshot since skip_bigwig=true is now applied via the prokaryotic config. Task count 58 -> 46. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

test: load prokaryotic config in nf-test instead of duplicating params

adamrtalbot

lgtm

The seqera conda channel only has ARM builds of STAR 2.6.1d. The default (non-ARM) configs were incorrectly using seqera::star=2.6.1d, which fails conda environment creation on x86_64 CI runners. Switch to bioconda::star=2.6.1d for the default configs, keeping the seqera channel override in conf/arm.config for ARM builds. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Fix iGenomes STAR conda channel for x86_64

gburnett-nvidia and others added 30 commits February 4, 2026 08:20

Adding tag to ignore GPU workflows on the CPU runner

3d55e3b

Update snaps

cb2a979

Adding cpu tags to all other tests

ca2c728

docs: Add changelog entry for topic-based version reporting migration

e00a783

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

update pipeline snaps

adfc874

fix: Update sentieon version 202503.01 -> 202503.02

c7bc3ea

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

chore: Update picard ARM container to 3.4.0

278f9df

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

chore: Add Matthias Zepper to credits for PR review

c01d5f8

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

style: Fix table formatting in usage.md

715816b

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Fix snap

34599fc

Merge pull request #1689 from nf-core/feature/modules-topics-update

90f32d3

feat: Migrate to topic-based version reporting

test: Update prepare_genome snapshot for topic-based versions

00d141e

Update snapshot after topic-based version reporting migration. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Update snap

28f1481

Merge pull request #1688 from nf-core/feature/prokaryotic-support

d20a221

feat: Add prokaryotic RNA-seq support with -profile prokaryotic

Merge remote-tracking branch 'origin/dev' into parabricks-unblock

5ed29f7

# Conflicts: # CHANGELOG.md # subworkflows/local/prepare_genome/main.nf

Add Parabricks GPU acceleration docs

fb6cfe6

Document the --use_parabricks_star flag in the usage guide, covering requirements, expected behaviour, and known limitations versus CPU STAR. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Update a couple of snaps

3e4be3f

fix: Inline variable declarations in nf-test.config for linter

415bc94

The Nextflow linter does not allow `def` variable declarations mixed with config statements in a config block. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Revert "fix: Inline variable declarations in nf-test.config for linter"

d36b186

This reverts commit 415bc94.

pinin4fjords and others added 2 commits February 27, 2026 11:42

docs: add PR #1730 to changelog

897bb4c

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

ci: bump minimum Nextflow version to 25.04.3

618a39b

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

pinin4fjords mentioned this pull request Feb 27, 2026

refactor: code quality improvements from PR #1717 review #1733

Closed

1 task

pinin4fjords and others added 2 commits February 27, 2026 11:46

Merge pull request #1730 from nf-core/fix/nxf-version-conda-mkdir

a0fafe0

ci: bump Nextflow to 25.04.3 to fix conda --mkdir bug

This was referenced Feb 27, 2026

refactor: compose test_prokaryotic profile on prokaryotic #1734

Merged

refactor: code quality improvements from PR #1717 review #1735

Merged

pinin4fjords and others added 10 commits February 27, 2026 12:02

Merge branch 'dev' into fix/prokaryotic-profile-composability

bd06b4a

docs: add PR #1735 to changelog

288e5fd

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

docs: add PR #1734 to changelog

c7ca5e5

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

merge: resolve conflicts with dev after #1730 merge

8a38278

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

ci: drop latest-everything from GPU workflow

b746123

GPU runners are expensive - just test minimum version. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

fix: align ch_transcript_fasta_placeholder assignment

4ce1a57

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

Merge pull request #1735 from nf-core/fix/pr1717-code-quality

15f75cf

refactor: code quality improvements from PR #1717 review

Merge branch 'dev' into fix/prokaryotic-profile-composability

b91f906

Merge pull request #1734 from nf-core/fix/prokaryotic-profile-composa…

12ca2e0

…bility refactor: compose test_prokaryotic profile on prokaryotic

pinin4fjords mentioned this pull request Feb 27, 2026

test: load prokaryotic config in nf-test instead of duplicating params #1736

Merged

2 tasks

docs: add PR #1736 to changelog

e3d5469

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

adamrtalbot reviewed Feb 27, 2026

View reviewed changes

Comment thread docs/usage.md

pinin4fjords and others added 3 commits February 27, 2026 13:11

Merge pull request #1736 from nf-core/fix/prokaryotic-nftest-use-profile

35ce3a7

test: load prokaryotic config in nf-test instead of duplicating params

pinin4fjords mentioned this pull request Feb 27, 2026

Reorganise usage docs with a user-centric structure #1737

Open

pinin4fjords requested a review from adamrtalbot February 27, 2026 13:38

adamrtalbot approved these changes Feb 27, 2026

View reviewed changes

pinin4fjords and others added 3 commits February 27, 2026 13:56

Merge branch 'dev' into fix/igenomes-star-conda

fdffb6f

Merge pull request #1738 from nf-core/fix/igenomes-star-conda

e8ba853

Fix iGenomes STAR conda channel for x86_64

pinin4fjords merged commit 13328f7 into master Feb 27, 2026
170 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Dev -> master for nf-core/rnaseq 3.23.0#1717

Dev -> master for nf-core/rnaseq 3.23.0#1717
pinin4fjords merged 394 commits into
masterfrom
dev

pinin4fjords commented Feb 26, 2026 •

edited

Loading

Uh oh!

Uh oh!

adamrtalbot left a comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

Conversation

pinin4fjords commented Feb 26, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Release 3.23.0

Review guide

New features

Bug fixes and improvements

Refactoring and maintenance

Documentation

CI fixes

New parameters

New profiles

Test plan

Uh oh!

Uh oh!

adamrtalbot left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

7 participants

pinin4fjords commented Feb 26, 2026 •

edited

Loading