RNA-seq Genome-Guided Assembly Pipeline (Nextflow)

This Nextflow DSL2 pipeline performs RNA-seq quality control, genome alignment, genome-guided de novo transcriptome assembly with Trinity, coding sequence prediction with TransDecoder, completeness assessment with BUSCO, protein similarity search with BLASTx, sample-level quantification with Salmon, and differential expression with DESeq2 (via tximport). Optional featureCounts is provided if a gene annotation is available.

Features

• FastQC per-sample QC and MultiQC aggregation
• HISAT2 genome index and paired-end alignment, sorted/indexed BAMs
• Trinity genome-guided assembly across all BAMs
• TransDecoder ORF prediction
• BUSCO transcriptome completeness
• BLASTx annotation against a reference protein FASTA
• Salmon quantification against assembled transcripts
• DESeq2 differential expression using tximport
• Optional featureCounts quantification (requires GTF/GFF)

Requirements

• Nextflow 22.10+ (nextflow -version)
• Either Conda/Mamba (recommended) or Docker/Singularity

Input samplesheet

CSV with header and four columns: sample,condition,fq1,fq2.

• sample: short sample identifier (no spaces)
• condition: group label (e.g., M for male, F for female)
• fq1: path to read 1 FASTQ (gzip accepted)
• fq2: path to read 2 FASTQ (gzip accepted)

Example:

sample,condition,fq1,fq2
F1,F,data/IT_F1_1.fq.gz,data/IT_F1_2.fq.gz
F2,F,data/IT_F2_1.fq.gz,data/IT_F2_2.fq.gz
F3,F,data/IT_F3_1.fq.gz,data/IT_F3_2.fq.gz
M1,M,data/IT_M1_1.fq.gz,data/IT_M1_2.fq.gz
M2,M,data/IT_M2_1.fq.gz,data/IT_M2_2.fq.gz
M3,M,data/IT_M3_1.fq.gz,data/IT_M3_2.fq.gz

Quick start

Run with Conda:

nextflow run main.nf -profile standard,conda \
  --genome data/genome.fasta \
  --samplesheet data/samplesheet.csv \
  --ref_proteins data/ref_proteins.fasta \
  --busco_db insecta_odb10 \
  --outdir results \
  --trinity_prefix gg_trinity \
  --max_intron 10000 \
  --de_case_label M --de_control_label F

Optional: featureCounts if you have a gene annotation:

nextflow run main.nf -profile standard,conda \
  --genome data/genome.fasta \
  --samplesheet data/samplesheet.csv \
  --ref_proteins data/ref_proteins.fasta \
  --busco_db insecta_odb10 \
  --featurecounts_gtf data/annotation.gtf

Outputs

• results/qc/fastqc/ — FastQC reports per sample
• results/qc/ — MultiQC report
• results/genome_index/ — HISAT2 genome index
• results/*.bam and *.bam.bai — Sorted and indexed BAMs per sample
• results/gg_trinity_assembly/Trinity-GG.fasta — Assembled transcripts
• results/gg_trinity_assembly/Trinity-GG.fasta.transdecoder.* — TransDecoder outputs
• results/busco/ — BUSCO results (busco_out/)
• results/blastdb/ — BLAST protein database
• results/blastx.outfmt6 — BLASTx tabular hits (outfmt 6)
• results/salmon_index/ — Salmon index of assembled transcripts
• results/salmon_quant/ — Per-sample quant.sf
• results/de/ — Differential expression: deseq2_results.tsv, normalized_counts.tsv, plots/
• results/counts.txt — featureCounts matrix (if --featurecounts_gtf is provided)

Parameters (selected)

• --genome (string): Path to genome FASTA
• --samplesheet (string): CSV with sample,condition,fq1,fq2
• --ref_proteins (string): Path to protein FASTA for BLASTx DB
• --busco_db (string): BUSCO lineage name (e.g. insecta_odb10) or local lineage path
• --outdir (string): Output directory (default: results)
• --trinity_prefix (string): Prefix for assembly output directory name
• --max_intron (int): Trinity --genome_guided_max_intron (default: 10000)
• --de_case_label (string): Case condition label (default: M)
• --de_control_label (string): Control condition label (default: F)
• --featurecounts_gtf (string|null): GTF/GFF path for featureCounts
• --max_cpus (int): Max CPUs per process (default: 16)
• --max_memory (string): Max memory per process (default: 100.GB)

Notes

• DE is performed at transcript-level using Salmon + tximport + DESeq2; adjust --de_case_label and --de_control_label to match your groups.
• BUSCO may download lineage data on first run; set --busco_db to a local path to avoid downloads.
• featureCounts requires a valid annotation (GTF/GFF). Counting against a FASTA is not supported.