[study] BERNARDINI2024_TIRE (PRJNA856813)

.github/workflows/ci.yml

@@ -198,6 +198,23 @@ jobs:
             --viral_reference data/reference/pathogens/OsHV-1/NC_005881.2.fasta \
             --outdir /tmp/aree-smoke/rnaseq-raw-se \
             -work-dir /tmp/aree-smoke/work-rnaseq-raw-se
+
+      - name: Exercise the RNA-seq raw DAG single-end, as 3' Tag-seq
+        run: |
+          nextflow run workflows/rnaseq -profile local,ci_stub -stub-run \
+            --mode raw_reanalysis \
+            --single_end --tag_seq \
+            --study_id CI_TAGSEQ \
+            --comparison_id treatment_vs_control \
+            --reads 'tests/fixtures/rnaseq_raw_se/ERR*.fastq' \
+            --transcript_fasta tests/fixtures/rnaseq_raw/reference.fa \
+            --gtf tests/fixtures/rnaseq_raw/annotation.gtf \
+            --tx2gene tests/fixtures/rnaseq_raw/tx2gene.tsv \
+            --sample_sheet tests/fixtures/rnaseq_raw_se/samples.tsv \
+            --control_level control \
+            --treatment_level treatment \
+            --outdir /tmp/aree-smoke/rnaseq-tagseq \
+            -work-dir /tmp/aree-smoke/work-rnaseq-tagseq
           # One trimmed file per library, none for a second mate.
           test "$(ls /tmp/aree-smoke/rnaseq-raw-se/rnaseq/trimmed/*_trimmed_R1.fastq.gz | wc -l)" -eq 4
           ! ls /tmp/aree-smoke/rnaseq-raw-se/rnaseq/trimmed/*_trimmed_R2.fastq.gz 2>/dev/null
@@ -211,6 +228,7 @@ jobs:
           test -f /tmp/aree-smoke/rnaseq-raw/rnaseq/standardized/CI_RNASEQ_treatment_vs_control_dge_standardized.tsv
           test -f /tmp/aree-smoke/rnaseq-raw/rnaseq/viral_load/CI_RNASEQ_treatment_vs_control_viral_load.tsv
           test -f /tmp/aree-smoke/rnaseq-raw-se/rnaseq/standardized/CI_RNASEQ_SE_treatment_vs_control_dge_standardized.tsv
+          test -f /tmp/aree-smoke/rnaseq-tagseq/rnaseq/standardized/CI_TAGSEQ_treatment_vs_control_dge_standardized.tsv
 
       # Each manifest's software_versions must match the versions.yml every
       # task of that run wrote, so it cannot fall back to declared pins.

config/BERNARDINI2024_TIRE.config

@@ -0,0 +1,77 @@
+/*
+ * AREE :: run configuration for BERNARDINI2024_TIRE (PRJNA856813)
+ *
+ * Bernardini et al. 2024, J Hazard Mater 480:135742
+ * (doi:10.1016/j.jhazmat.2024.135742). Adult oysters exposed to tire
+ * microparticles (MR) or their leachate (LEA), each at a low (L) and high (H)
+ * concentration, vs unexposed controls (CTL); gill sampled on day 41, one
+ * oyster per library, 6 per group. Single-end 76 bp 3' QuantSeq-style
+ * libraries (poly-A content 7.4% of reads vs 0.1% for full-length RNA-seq,
+ * checked 2026-10-12), so --single_end --tag_seq.
+ *
+ * Four prespecified comparisons against the day-41 gill controls; default =
+ * the meta_analysis_primary one (tire microparticles, high concentration):
+ *
+ *   nextflow run workflows/rnaseq -config config/BERNARDINI2024_TIRE.config \
+ *     -profile local,workstation
+ *
+ *   --comparison_id tire_particles_low_vs_ctl_d41_gill  --treatment_level NovaSeq_6000_D41_GI_MR_L
+ *   --comparison_id tire_leachate_high_vs_ctl_d41_gill  --treatment_level NovaSeq_6000_D41_GI_LEA_H
+ *   --comparison_id tire_leachate_low_vs_ctl_d41_gill   --treatment_level NovaSeq_6000_D41_GI_LEA_L
+ *
+ * The BioSamples are 16S "metagenome" records shared by each oyster's 16S
+ * (MiSeq) and RNA-seq (NovaSeq) runs, so the design is parsed from the
+ * run-level experiment_title, platform included (only NovaSeq runs are used):
+ *
+ *   aree fetch-samplesheet --bioproject PRJNA856813 --study BERNARDINI2024_TIRE \
+ *     --label-field experiment_title \
+ *     --label-pattern '(?P<platform>NovaSeq 6000|MiSeq) sequencing: (?P<day>D41|D44|T0)_(?P<compartment>GI_|DG_|FSW_|)(?P<treatment>CTL_|NAT_|MR_L_|MR_H_|LEA_L_|LEA_H_|T0_|)(?P<replicate>[1-6])$' \
+ *     --condition-attribute platform --condition-attribute day \
+ *     --condition-attribute compartment --condition-attribute treatment \
+ *     --attribute replicate --attribute ref_biomaterial --attribute collection_date
+ */
+
+includeConfig 'base.config'
+
+params {
+    mode = 'raw_reanalysis'
+
+    rnaseq {
+        study_id           = 'BERNARDINI2024_TIRE'
+        // Null so --comparison_id on the command line takes effect.
+        comparison_id      = null
+        genome_assembly    = 'GCF_963853765.1'
+        annotation_version = 'GCF_963853765.1-RS_2024_06'
+        processed_results  = null
+    }
+
+    comparison_id = 'tire_particles_high_vs_ctl_d41_gill'
+
+    sample_sheet = 'data/studies/BERNARDINI2024_TIRE/samplesheet.tsv'
+
+    control_level   = 'NovaSeq_6000_D41_GI_CTL'
+    treatment_level = 'NovaSeq_6000_D41_GI_MR_H'
+
+    single_end = true
+    tag_seq    = true
+
+    // Only the 30 day-41 gill RNA-seq runs (CTL, MR_L, MR_H, LEA_L, LEA_H) are
+    // staged (~8.6 GB); see the study YAML. Verify against ENA's MD5s from
+    // data/studies/BERNARDINI2024_TIRE/fastq_manifest.tsv.
+    reads = 'data/raw/BERNARDINI2024_TIRE/SRR*.fastq.gz'
+
+    transcript_fasta = 'data/reference/GCF_963853765.1/rna.fna'
+    gtf              = 'data/reference/GCF_963853765.1/genomic.gtf'
+    tx2gene          = 'data/reference/GCF_963853765.1/tx2gene.tsv'
+    salmon_index     = null
+
+    outdir = 'results/rnaseq/BERNARDINI2024_TIRE'
+}
+
+process {
+    // 30 NovaSeq 6000 single-end 76 bp libraries, 5.4-8.3 M reads. Sized for a 16 GB machine.
+    withName: 'SALMON_INDEX' { cpus = 8;  memory = '12 GB'; time = '4h'  }
+    withName: 'SALMON_QUANT' { cpus = 4;  memory = '8 GB';  time = '1h'  }
+    withName: 'TRIM_FASTP'   { cpus = 2;  memory = '4 GB';  time = '1h'  }
+    withName: 'FASTQC'       { cpus = 1;  memory = '2 GB';  time = '1h'  }
+}