Add RNA-seq Variant Discovery (RNAvar) workflow

workflows/variant-calling/rna-variant-discovery/.dockstore.yml

@@ -0,0 +1,11 @@
+version: 1.2
+workflows:
+- name: rna-variant-discovery
+  subclass: Galaxy
+  publish: true
+  primaryDescriptorPath: /RNAvar.ga
+  testParameterFiles:
+  - /RNAvar-tests.yml
+  authors:
+  - name: "Emre Gülben"
+    orcid: "https://orcid.org/0009-0009-9085-1055"

workflows/variant-calling/rna-variant-discovery/CHANGELOG.md

@@ -0,0 +1,5 @@
+# Changelog
+
+## [0.1] - 2026-03-27
+
+- First release.

workflows/variant-calling/rna-variant-discovery/README.md

@@ -0,0 +1,37 @@
+# RNA-seq Variant Discovery (RNAvar)
+
+This workflow is a faithful reimplementation of the [nf-core/rnavar](https://nf-co.re/rnavar) pipeline for the Galaxy ecosystem. It handles the transition from raw reads to annotated variants using industry-standard tools and follows the established best practices for RNA-seq variant discovery.
+
+## Workflow Logic
+1. **Alignment:** **STAR** in 2-pass mode for high-accuracy splice-aware mapping.
+2. **Preprocessing:** **MarkDuplicates** and **GATK4 SplitNCigarReads** to prepare the BAM for variant calling by splitting reads into exon segments.
+3. **Recalibration:** **GATK4 BaseRecalibrator** (BQSR) using known polymorphic sites to adjust base quality scores.
+4. **Variant Calling:** **GATK4 HaplotypeCaller** with parameters specifically tuned for RNA-seq (e.g., ignoring soft-clipped bases).
+5. **Filtering:** **bcftools filter** applying hard filters (`FS > 30.0`, `QD < 2.0`) to reduce false positives.
+6. **Annotation:** Functional annotation of variants using **SnpEff** against a selected local database.
+
+## Inputs
+* **Forward Reads (R1)**: Fastq sequencing reads (forward). Supports gzipped (`.gz`) files.
+* **Reverse Reads (R2)**: Fastq sequencing reads (reverse). Supports gzipped (`.gz`) files.
+* **Reference Genome**: FASTA file of the reference genome (e.g., hg38).
+* **Genome Annotation**: GFF3 file containing gene models for splice-aware alignment.
+* **Known Variants (dbSNP)**: VCF file of known polymorphisms for base quality score recalibration.
+* **Known Indels**: VCF file of known insertions and deletions for base quality score recalibration.
+* **SnpEff Genome Version**: The name of the locally installed SnpEff database (e.g., hg38).
+* **Variant Exclusion Filter**: A bcftools-style expression for hard-filtering variants (default: FS > 30.0 || QD < 2.0).
+* **STAR Genome Overhang**: Sequence length around junctions for STAR indexing. Set to (Read Length - 1). For example, use 99 for 100bp reads.
+
+## Outputs
+* **Final Annotated Variants**: VCF file containing discovered variants with functional annotations (impact, effect, gene names) from SnpEff.
+* **MultiQC Quality Report**: An HTML report aggregating quality metrics from alignment and preprocessing steps.
+
+## Testing and Parity
+The workflow is validated against a targeted subset of human RNA-seq data on chromosome 22, sourced from the [nf-core/rnavar test dataset](https://github.com/nf-core/test-datasets/tree/rnavar). It achieves technical parity with the reference discovery engine, identifying 54 high-confidence variants.
+
+## Limitations
+Compared to the full [nf-core/rnavar (v1.2.3)](https://nf-co.re/rnavar/1.2.3/docs/usage/) implementation, this Galaxy workflow currently has the following limitations:
+* **Sample Handling:** Does not support CSV sample sheets; inputs must be provided as individual datasets or collections.
+* **Modularity:** Base recalibration (BQSR) and variant filtering are mandatory steps, whereas they are optional in the original pipeline.
+* **Preprocessing:** No current support for UMI-Tools-based deduplication or optional read trimming.
+* **Annotation:** Variant annotation is restricted to SnpEff; support for other alternatives is not yet implemented.
+* **Performance:** The `GATK4 SplitNCigarReads` step is not parallelized, which may impact runtime on very large datasets compared to the Nextflow execution model.

workflows/variant-calling/rna-variant-discovery/RNAvar-tests.yml

@@ -0,0 +1,46 @@
+- doc: Test outline for RNAvar
+  job:
+    "Forward Reads (R1)":
+      class: File
+      path: test-data/test_rnaseq_1.fastq.gz
+      filetype: fastqsanger.gz
+    "Reverse Reads (R2)":
+      class: File
+      path: test-data/test_rnaseq_2.fastq.gz
+      filetype: fastqsanger.gz
+    "Reference Genome":
+      class: File
+      path: test-data/genome.fasta
+      filetype: fasta
+    "Genome Annotation":
+      class: File
+      path: test-data/genome.gff3
+      filetype: gff3
+    "Known Variants (dbSNP)":
+      class: File
+      path: test-data/dbsnp_146.hg38.vcf
+      filetype: vcf
+    "Known Indels":
+      class: File
+      path: test-data/mills_and_1000G.indels.vcf
+      filetype: vcf
+    "SnpEff Genome Version": hg38
+    "Variant Exclusion Filter": "FS > 30.0 || QD < 2.0"
+    "STAR Genome Overhang": 14
+  outputs:
+    "Final Annotated Variants":
+      asserts:
+        # Expected variants: 54 | VCF Header lines: 36 | Total: 90.
+        - has_n_lines:
+            n: 90
+        # Verify the content is scientifically correct by checking for 
+        # the presence of the target chromosome string.
+        - has_text:
+            text: "chr22"
+        # Ensure SnpEff successfully added functional annotations
+        - has_text:
+            text: "ANN="
+    "MultiQC Quality Report":
+      asserts:
+        - has_text:
+            text: "MultiQC"