SomaticWrapper v3.0.0 (Compute1)

Automated Somatic Variant Calling Pipeline (HG38)

SomaticWrapper is a fully automated and modular pipeline for detecting somatic variants from paired tumor–normal WGS/WXS data on the LSF compute1 cluster (WashU).
It integrates multiple industry-standard variant callers — Strelka2, VarScan2, Mutect1, and Pindel — and produces comprehensive, annotated mutation calls in MAF format.

---

🔬 Overview

• SNV calls: intersection of 2 out of 3 callers — Strelka2, Mutect1, VarScan2
• Indel calls: intersection of 2 out of 3 callers — Strelka2, VarScan2, Pindel
• Reference genome: Human GRCh38 (HG38)
• Scheduler: LSF (supports job dependencies and groups)

Final output files:

• dnp.annotated.maf → all variants
• dnp.annotated.coding.maf → coding variants only

---

🚀 Improvements (v3.0.0)

1. Added Step 0 — automatically submits the full pipeline (Steps 1 → 11) with job dependencies (j2 waits for j1, etc.).

2. Added Step 22 — automatically submits the full pipeline (Steps 2 → 11) with job dependencies (j3 waits for j2, etc.).

3. Added Step 23 — automatically submits the full pipeline (Steps 3 → 11) with job dependencies

4. Added Step 24 — automatically submits the full pipeline (Steps 4 → 11) with job dependencies

5. Added Step 25 — automatically submits the full pipeline (Steps 5 → 11) with job dependencies

6. Added Step 26 — automatically submits the full pipeline (Steps 6 → 11) with job dependencies

7. Added Step 27 — automatically submits the full pipeline (Steps 7 → 11) with job dependencies

8. Added Step 28 — automatically submits the full pipeline (Steps 8 → 11) with job dependencies

9. Added Step 29 — automatically submits the full pipeline (Steps 9 → 11) with job dependencies

10. Added Step 30 — automatically submits the full pipeline (Steps 10 → 11) with job dependencies

---

⚙️ Environment Setup (Compute1)

Before running, update your ~/.bashrc to include the necessary environment variables:

export PATH=/storage1/fs1/songcao/Active/Software/anaconda3/bin:$PATH
export STORAGE1=/storage1/fs1/songcao/Active
export STORAGE2=/storage1/fs1/dinglab/Active
export STORAGE3=/storage1/fs1/m.wyczalkowski/Active
export LSF_DOCKER_VOLUMES="$STORAGE1:$STORAGE1 $STORAGE2:$STORAGE2 $STORAGE3:$STORAGE3"

Then activate:

source ~/.bashrc

---

🧩 Usage

Step 1. Download or clone this repository

git clone https://github.com/YourGitRepo/somaticwrapper.git
cd somaticwrapper

Step 2. Prepare your run and log directories

Example:

mkdir -p /storage1/fs1/songcao/Active/Projects/somatic/example_run_somatic_2025
mkdir -p /storage1/fs1/songcao/Active/Projects/somatic/example_run_somatic_2025/log

Step 3. Run the pipeline

Option A – Run all steps automatically (recommended)

Use --step 0 to run Steps 1–14 sequentially with built-in job dependencies:

perl somaticwrapper.pl   --step 0   --rdir /storage1/fs1/songcao/Active/Projects/somatic/example_run_somatic_2025   --log  /storage1/fs1/songcao/Active/Projects/somatic/example_run_somatic_2025/log   --ref /storage1/fs1/songcao/Active/Database/hg38_database/GRCh38.d1.vd1/GRCh38.d1.vd1.fa   --smg /storage1/fs1/songcao/Active/Database/SMG/smg_list.txt   --groupname example_run_somatic_2025   --users scao   --wgs 0   --srg 1   --sre 0   --exonic 1   --q long   --mincovt 14 --mincovn 8 --minvaf 0.05 --maxindsize 100

Option B – Run a specific step manually

perl somaticwrapper.pl --step 5 --rdir <run_dir> --log <log_dir> ...

---

🔢 Step Reference

Step	Description
0	Submit steps (1–11) automatically with dependencies
1	Run Strelka2
2	Run VarScan2
3	Run Pindel
4	Run Mutect1
5	Parse Mutect results
6	Parse Strelka2 results
7	Parse VarScan2 results
8	Parse Pindel results
9	QC VCF files
10	Merge VCF files
11	Generate MAF files
12	Merge run-level MAF
13	DNP annotation
14	Clean unnecessary intermediate files
22	Submit steps (2–11) automatically with dependencies
23	Submit steps (3–11) automatically with dependencies
24	Submit steps (4–11) automatically with dependencies
25	Submit steps (5–11) automatically with dependencies
26	Submit steps (6–11) automatically with dependencies
27	Submit steps (7–11) automatically with dependencies
28	Submit steps (8–11) automatically with dependencies
29	Submit steps (9–11) automatically with dependencies
30	Submit steps (10–11) automatically with dependencies

---

⚙️ Key Parameters

Parameter	Description
--rdir	Full path to run directory containing per-sample folders
--log	Path for log output (usually parent of rdir)
--srg	BAM has read groups (1 = yes, 0 = no)
--sre	Rerun and overwrite results (1 = yes, 0 = no)
--wgs	1 = WGS, 0 = WXS
--groupname	Job group name
--users	LSF user account (used in job group path)
--ref	HG38 reference FASTA
--smg	SMG gene list file
--q	LSF queue (research-hpc, ding-lab, or long)
--mincovt	Minimum tumor coverage (≥ 14)
--mincovn	Minimum normal coverage (≥ 8)
--minvaf	Minimum variant allele frequency (≥ 0.05)
--maxindsize	Maximum indel size (≤ 100)
--exonic	Output exonic region (1 = yes, 0 = no)

---

🧾 Example Output Files

run_dir/
├── <sample_name>/
│   ├── strelka/
│   ├── varscan/
│   ├── pindel/
│   ├── mutect1/
│   ├── merged.withmutect.vcf
│   ├── <sample>.withmutect.maf
│   └── <sample>.dnp.annotated.maf
└── log/
    ├── LSF_DIR_SOMATIC/
    └── tmpsomatic/

---

👤 Contact

Author: Song Cao
Email: scao@wustl.edu
Washington University in St. Louis

---