WGS Workflow — BioHPC Cornell

Nextflow DSL2 pipeline for whole-genome sequencing analysis on BioHPC Cornell servers.

Modules: fastp → bowtie2 → samtools → picard (MarkDuplicates) → qualimap → deeptools (GC bias) → MultiQC → Quarto GC bias report

---

Prerequisites (BioHPC)

• Nextflow ≥ 23.x (module load nextflow or available in your PATH)
• Apptainer/Singularity — available by default on BioHPC nodes (Docker is not required)
• No local software installation required — all tools run inside Singularity containers pulled automatically from GHCR

---

Quick Start

nextflow run bixBeta/wgs -r container \
  --id PROJECT_ID \
  --sheet sample-sheet.csv \
  --genome hg38 \
  --bowtie2 \
  --fastp

---

Sample Sheet

Tab or comma-separated file with three columns:

label,fastq1,fastq2
SAMPLE1,/path/to/SAMPLE1_R1.fastq.gz,/path/to/SAMPLE1_R2.fastq.gz
SAMPLE2,/path/to/SAMPLE2_R1.fastq.gz,/path/to/SAMPLE2_R2.fastq.gz

If your FASTQs are in a fastqs/ subdirectory in the run folder, pass --fastqs and use just the filenames (no path) in the sheet.

---

Parameters

Parameter	Default	Description
--id	TREx_ID	Project ID — used for output file naming
--sheet	sample-sheet.csv	Path to sample sheet
--genome	null	Reference genome (see below)
--mode	PE	Sequencing mode: PE (paired-end)
--bowtie2	false	Enable bowtie2 alignment
--fastp	false	Enable fastp trimming (recommended)
--gcbias	false	Enable GC bias analysis (requires --bowtie2)
--fastqs	false	Look for FASTQs in ./fastqs/ subdirectory
--listGenomes	false	Print available genome index paths and exit

---

Available Genomes

Bowtie2 indices live at /workdir/genomes/ on the BioHPC server.

Key	Species	Assembly
hg38	Human	GRCh38
mm10	Mouse	GRCm38
dm6	Drosophila	BDGP6
canFam4	Dog	canFam4
fc9	Cat	Felis_catus_9.0
vitis	Grapevine	GCA_000003745.2

Run --listGenomes to see full index paths.

---

Example Commands

Standard run (trim + align + QC):

nextflow run bixBeta/wgs -r container \
  --id 7054D \
  --sheet sample-sheet.csv \
  --genome hg38 \
  --fastp \
  --bowtie2

With GC bias analysis:

nextflow run bixBeta/wgs -r container \
  --id 7054D \
  --sheet sample-sheet.csv \
  --genome mm10 \
  --fastp \
  --bowtie2 \
  --gcbias

FASTQs in a subdirectory:

nextflow run bixBeta/wgs -r container \
  --id 7054D \
  --sheet sample-sheet.csv \
  --genome hg38 \
  --fastp \
  --bowtie2 \
  --fastqs

List available genomes:

nextflow run bixBeta/wgs -r container --listGenomes

---

Containers

All images are hosted on GHCR and pulled automatically by Nextflow.

Image	Tools
ghcr.io/bixbeta/wgs-align:latest	fastp, bowtie2, samtools, picard
ghcr.io/bixbeta/wgs-qualimap:latest	qualimap
ghcr.io/bixbeta/wgs-deeptools:latest	deeptools
ghcr.io/bixbeta/wgs-quarto:latest	R 4.4.2, Quarto 1.5.57
multiqc/multiqc:v1.32	MultiQC

---

Output Structure

./
├── trimmed_fastqs/          # fastp trimmed reads
├── trimmed_logs/            # fastp HTML + JSON reports
├── primary_BAMS/            # bowtie2 aligned BAMs
├── DEDUP_BAMS/              # deduplicated BAMs
├── STATS/
│   ├── SAMTOOLS/            # flagstat + idxstats
│   ├── PICARD/              # MarkDuplicates metrics
│   └── QUALIMAP_RES/        # bamqc output
├── GCBias_DeepTools/        # GC bias PNGs + TXT (if --gcbias)
└── Reports/
    ├── <genome>/            # MultiQC HTML report
    └── <genome>/gc_bias_report.html  # Quarto GC bias report (if --gcbias)

---

Tips for BioHPC

• Run inside a screen or tmux session so the job persists if your SSH connection drops
• Submit via a SLURM wrapper to run Nextflow itself on a compute node:

  srun --pty --mem=8G --cpus-per-task=2 bash
  nextflow run bixBeta/wgs -r container ...

• Use -resume to restart from the last successful step after a failure:

  nextflow run bixBeta/wgs -r container ... -resume

• Nextflow work files are written to ./work/ — clean up with nextflow clean -f after a successful run