Add rpbp test data: per-stage intermediates for nf-core/modules#11695

README.md

@@ -464,6 +464,17 @@ The earth sciences folder contain subfolders for different data formats encounte
         - Homo_sapiens.GRCh38.111_chr19_22.pc.gtf.gz: chr19+chr22 of Ensembl 111 GTF, subset to `gene_biotype "protein_coding"` with lean attributes
         - bams/SRX1178088{5,6,7,8}.chr19_22.ds50.bam(+ .bai): 4-sample Ribo-seq cohort (same upstream SRX accessions as the chr20 BAMs above) at 50% downsample, filtered to chr19+chr22 and to reads overlapping protein-coding gene loci. 4 samples is the empirical PRICE-cohort floor; 3 samples crashes its noise-model inference.
         - README.md: full derivation recipe, empirical justification for the chosen subset/cohort, and the PRICE invocation used for verification.
+      - rpbp
+        - reference.annotated.bed.gz: transcript-level annotated BED output by rpbp/preparegenome, for testing rpbp/extractmetageneprofiles
+        - reference.orfs-genomic.annotated.bed.gz: genomic-coordinate ORF BED output by rpbp/preparegenome, for testing rpbp/extractorfprofiles and rpbp/estimateorfbayesfactors
+        - reference.orfs-exons.annotated.bed.gz: exon-coordinate ORF BED output by rpbp/preparegenome, for testing rpbp/extractorfprofiles
+        - SRX11780888_chr20.metagene-profile.csv.gz: metagene profile output by rpbp/extractmetageneprofiles, for testing rpbp/estimatemetagenebayesfactors
+        - SRX11780888_chr20.metagene-periodicity-bayes-factors.csv.gz: per-length metagene Bayes-factor table output by rpbp/estimatemetagenebayesfactors, for testing rpbp/selectperiodicoffsets
+        - SRX11780888_chr20.periodic-offsets.csv.gz: per-length periodic-offset table output by rpbp/selectperiodicoffsets, for testing rpbp/getperiodiclengthsoffsets
+        - SRX11780888_chr20.periodic_lengths_offsets.tsv: filtered length/offset pairs (lenient thresholds for chr20) output by rpbp/getperiodiclengthsoffsets, for testing rpbp/extractorfprofiles
+        - SRX11780888_chr20.profiles.mtx.gz: per-ORF Ribo-seq read-count profile matrix output by rpbp/extractorfprofiles, for testing rpbp/estimateorfbayesfactors
+        - SRX11780888_chr20.bayes-factors.bed.gz: per-ORF Bayes-factor table output by rpbp/estimateorfbayesfactors, for testing rpbp/selectfinalpredictionset
+        - README.md: per-file derivation recipe.
       - ribocode
         - genome_updated.gtf.gz: GTF with gene names updated via ribocode/gtfupdate, compressed for efficient storage
         - annotation.tar.gz: Tarball containing annotation directory output from ribocode/prepare for testing ribocode/metaplots and ribocode/ribocode modules

data/genomics/homo_sapiens/riboseq_expression/rpbp/README.md

@@ -0,0 +1,56 @@
+# Test data for `rpbp/*` modules
+
+Per-stage intermediates from a single end-to-end `fasta_gtf_bam_rpbp` subworkflow run on the existing chr20 fixture (`SRX11780888_chr20.bam` + `Homo_sapiens.GRCh38.111_chr20.gtf`). Each fixture is the immediate-upstream input for one rpbp module, so module-level tests can fetch one static file rather than chain six upstream stages.
+
+## Why these fixtures?
+
+`modules/nf-core/rpbp/*` test setups used to chain `GUNZIP -> PREPAREGENOME -> EXTRACTMETAGENEPROFILES -> ESTIMATEMETAGENEBAYESFACTORS -> SELECTPERIODICOFFSETS -> GETPERIODICLENGTHSOFFSETS -> EXTRACTORFPROFILES -> ESTIMATEORFBAYESFACTORS` to test downstream modules like `SELECTFINALPREDICTIONSET`. Every chained run cost several minutes of CI time per module. With these fixtures, each module test fetches its one immediate-upstream output and runs in well under a minute.
+
+The end-to-end integration test still lives in `subworkflows/nf-core/fasta_gtf_bam_rpbp/tests/main.nf.test`, which exercises the full chain.
+
+## Files
+
+| File | Size | Description |
+|---|---|---|
+| `reference.annotated.bed.gz` | <500 KB | Transcript-level annotated BED output by `rpbp/preparegenome`. Consumed by `rpbp/extractmetageneprofiles`. |
+| `reference.orfs-genomic.annotated.bed.gz` | <500 KB | Genomic-coordinate ORF BED output by `rpbp/preparegenome`. Consumed by `rpbp/extractorfprofiles` and `rpbp/estimateorfbayesfactors`. |
+| `reference.orfs-exons.annotated.bed.gz` | <500 KB | Exon-coordinate ORF BED output by `rpbp/preparegenome`. Consumed by `rpbp/extractorfprofiles`. |
+| `SRX11780888_chr20.metagene-profile.csv.gz` | <50 KB | Per-read-length metagene profile output by `rpbp/extractmetageneprofiles`. Consumed by `rpbp/estimatemetagenebayesfactors`. |
+| `SRX11780888_chr20.metagene-periodicity-bayes-factors.csv.gz` | <50 KB | Per-read-length periodicity Bayes-factor table output by `rpbp/estimatemetagenebayesfactors`. Consumed by `rpbp/selectperiodicoffsets`. |
+| `SRX11780888_chr20.periodic-offsets.csv.gz` | <50 KB | Per-read-length periodic-offset table output by `rpbp/selectperiodicoffsets`. Consumed by `rpbp/getperiodiclengthsoffsets`. |
+| `SRX11780888_chr20.periodic_lengths_offsets.tsv` | <1 KB | Filtered length/offset pairs output by `rpbp/getperiodiclengthsoffsets` using lenient `'10 1 None 0.0'` thresholds (chr20 alone does not pass the rpbp defaults). Consumed by `rpbp/extractorfprofiles`. |
+| `SRX11780888_chr20.profiles.mtx.gz` | <2 MB | Per-ORF Ribo-seq read-count profile sparse matrix output by `rpbp/extractorfprofiles`. Consumed by `rpbp/estimateorfbayesfactors`. |
+| `SRX11780888_chr20.bayes-factors.bed.gz` | <2 MB | Per-ORF translation-Bayes-factor table output by `rpbp/estimateorfbayesfactors`. Consumed by `rpbp/selectfinalpredictionset`. |
+
+All files <4 MB. Total set <10 MB.
+
+## How they were derived
+
+A single `nf-core subworkflows test fasta_gtf_bam_rpbp` run, with input
+
+- BAM: `aligned_reads/SRX11780888_chr20.bam` (+ `.bai`)
+- FASTA: `Homo_sapiens.GRCh38.dna.chromosome.20.fa.gz` (gunzipped)
+- GTF: `Homo_sapiens.GRCh38.111_chr20.gtf`
+
+from this same `riboseq_expression/` folder, plus the module-level test config:
+
+```
+process {
+    withName: 'RPBP_GETPERIODICLENGTHSOFFSETS' {
+        ext.args = '10 1 None 0.0'
+    }
+    withName: 'RPBP_SELECTFINALPREDICTIONSET' {
+        ext.args = '--select-longest-by-stop --select-best-overlapping'
+    }
+}
+```
+
+The `RPBP_GETPERIODICLENGTHSOFFSETS` thresholds (`min_metagene_profile_count=10`, `min_metagene_bf_mean=1`, `max_metagene_bf_var=None`, `min_metagene_bf_likelihood=0.0`) are intentionally lenient so chr20 produces non-empty output; rpbp's production defaults are tuned for whole-genome data and reject chr20-only profiles.
+
+Output files were captured directly from each module's work directory.
+
+## rpbp version
+
+`rpbp 4.0.1` (Wave container `community.wave.seqera.io/library/rpbp:4.0.1--71297b462026e13b`).
+
+Used by `modules/nf-core/rpbp/*/tests/main.nf.test`.

data/genomics/homo_sapiens/riboseq_expression/rpbp/SRX11780888_chr20.periodic_lengths_offsets.tsv

@@ -0,0 +1,10 @@
+length	offset
+22	11
+24	12
+25	12
+26	12
+27	12
+28	12
+29	12
+30	13
+33	13