PanVariants Usage Documentation

Overview

PanVariants is a highly accurate variant detection pipeline that integrates pangenome references and artificial intelligence (AI) model to identify single nucleotide variants (SNVs), INDELs (<50 bp), copy number variants (CNVs), short tandem repeats (STRs), and other structural variants (SVs, ≥50 bp). It supports execution on both local machines and Sun Grid Engine (SGE) clusters.

---

Quick Start

Simple Run Command

python3 PanVariant.py --samplelist samplelist --executor local

---

Install

git clone https://github.com/MGI-EU/PanVariants.git

Database download

• Here we take the download of databases related to the GRCh38 reference genome as an example; the download process for other reference genomes is similar.

sh PanVariants_required_files.sh

singularity .sif files download

• Download all .sif files from this URL: https://zenodo.org/records/19848582 and place them in the "sifs" folder.

Required Software

Software	Version	Installation
Nextflow	23.10.0	Download from: https://github.com/nextflow-io/nextflow/releases/download/v23.10.0/nextflow-23.10.0-all Rename to nextflow and copy to software directory
Java	≥ 17	Required for Nextflow execution
Singularity	≥ 3.8	Required for containerized execution
Python	≥ 3.9.13	Required for running the Python wrapper script

Optional Software

Software	Description
SGE (Sun Grid Engine)	Optional cluster management system for distributed computing (required only for blc executor)

---

Arguments Description

Required Arguments

Argument	Short	Description
--samplelist	-s	Sample information file. Must contain a single line with 3 tab-separated columns:sample_name   read1_path   read2_path
--executor	-ex	Execution engine options:- blc: Run pipeline using a Sun Grid Engine cluster- local: Run pipeline using the local machine

---

Optional Arguments

Argument	Short	Default	Description
--output	-o	result	Output directory path
--reference	-ref	GRCh38	Reference genome version. Options: GRCh38 or CHM13
--run_cnv	-rc	yes	Enable CNV detection (yes / no)
--run_sv	-rs	yes	Enable SV detection (yes / no)
--run_str	-rstr	yes	Enable STR detection (yes / no)
--mark_dup	-md	no	Mark duplicate reads (yes / no)
--fq_filter	-ft	no	Run FastQ filtering (yes / no)
--split_bam	-sb	no	Split BAM files by chromosome and run DeepVariant per chromosome (yes / no)
--adapter1	-ad1	AAGTCGGAGGCCAAGCGGTCTTAGGAAGACAA	Adapter 1 sequence
--adapter2	-ad2	AAGTCGGATCGTAGCCATGTCGTTCTGTGAGCCAAGGAGTTG	Adapter 2 sequence
--re_alignment	-ra	no	Perform re-alignment (yes / no)
--nf_config	-c	-	Path to Nextflow configuration file
--db	-d	-	Database path
--script	-sp	-	Script path
--sif	-f	-	Path to Singularity container image (.sif)
--scratch_tmp	-scr	-	Path for Nextflow scratch directive
--queue	-q	-	Queue name for the qsub -q parameter. Note: Only effective when --executor is set to blc.
--project	-pj	-	Project name for the qsub -P parameter. Note: Only effective when --executor is set to blc.
--ansi_log	-an	false	Enable ANSI logging for Nextflow (true / false)
--help	-h	-	Show help message and exit

---

Example Commands

Run Locally with All Variant Types Enabled

python3 PanVariant.py \
  --samplelist samples.tsv \
  --executor local \
  --output results \
  --reference GRCh38 \
  --run_cnv yes \
  --run_sv yes \
  --run_str yes

Run on SGE Cluster with Specific Queue and Project

python3 PanVariant.py \
  --samplelist samples.tsv \
  --executor blc \
  --queue high_mem \
  --project my_project \
  --output cluster_results

---

Notes

• The --queue and --project parameters are only valid when --executor is set to blc.
• The sample list file must strictly follow the format: three tab-separated columns with no header row.
• If using Singularity containers, ensure the correct .sif file path is provided via -f.
• Duplicate marking and FastQ filtering are disabled by default. Set them explicitly to yes if required.

---