Open-source bioinformatics benchmark tool — like Geekbench, but purpose-built for bioinformatics workloads.
BioMark simulates real bioinformatics computational workloads without requiring any tools to be installed. It benchmarks your machine's CPU, RAM, and SSD performance specifically for bioinformatics pipelines — giving you an honest assessment of what your hardware can and cannot do.
Designed for:
- 🎓 Students choosing a laptop for bioinformatics
- 🔬 Researchers evaluating hardware upgrades
- 💻 Lab managers planning computing infrastructure
- 📊 Anyone comparing machines for genomics work
Current version: v0.5.0 (Active Development)
BioMark is functional and actively developed. All 8 benchmark modules are working. The scoring system, HTML report, and hardware assessment are stable.
What works now:
- ✅ All 8 pipeline modules (DNA, RNA, scRNA, Protein, Epigenomics, Metagenomics, LongRead, Assembly)
- ✅ Automatic hardware detection (RAM, CPU, SSD, GPU)
- ✅ Honest PASS/FAIL scoring per pipeline step
- ✅ HTML report with charts and recommendations
- ✅ JSON export for programmatic use
Roadmap:
- 🔄 Community leaderboard website (biomark scores by machine)
- 🔄 Real dataset benchmarks (NA12878 WES reference sample)
- 🔄 Mac App Store release
- 🔄 Windows native support (currently via WSL2)
- 🔄 Tool paper submission
| Platform | Status | Notes |
|---|---|---|
| macOS (Apple Silicon M1-M5) | ✅ Fully supported | Recommended — tested on M1 Air |
| macOS (Intel) | ✅ Supported | Not optimized for Intel GPU |
| Linux (Ubuntu, Debian, CentOS) | ✅ Fully supported | Ideal for HPC nodes |
| Windows (WSL2) | Run inside Ubuntu WSL2 terminal | |
| Windows (native) | 🔄 Planned | Not yet supported natively |
| Module | Flag | Tools Simulated | Key Bottleneck |
|---|---|---|---|
| 🧬 DNA / WES | --dna |
BWA-MEM2, GATK, bcftools, FreeBayes, ANNOVAR | RAM (16GB+ for hg38) |
| 🔬 Bulk RNA-seq | --rna |
STAR, HISAT2, DESeq2, edgeR, clusterProfiler | RAM (32GB for STAR) |
| 🔬 scRNA-seq | --scrna |
Cell Ranger, Seurat, Scanpy, Monocle3 | RAM (32-64GB+) |
| 🔮 Protein Structure | --protein |
AlphaFold2, ESMFold, BLAST, FoldX | RAM + GPU |
| 🧬 Epigenomics | --epigenomics |
Bowtie2, MACS3, HOMER, DiffBind, chromVAR | RAM (16GB+) |
| 🦠 Metagenomics | --metagenomics |
Kraken2, MetaPhlAn4, QIIME2, HUMAnN3 | RAM (16GB+) + SSD |
| 📡 Long Read | --longread |
Dorado, Minimap2, Sniffles2, Medaka | RAM (16GB+) |
| 🔨 Assembly | --assembly |
SPAdes, Flye, Hifiasm, QUAST, BUSCO | RAM (32-500GB+) |
- Python 3.8+
- 4GB RAM minimum (to run BioMark itself)
- Works on macOS, Linux, and Windows WSL2
# Clone the repository
git clone https://github.com/shahbazigenomics/BioMark.git
cd BioMark
# Install dependencies
pip install -r requirements.txt# Show your system hardware profile
python src/main.py --sysinfo
# Run a specific module
python src/main.py --dna # DNA/WES pipeline
python src/main.py --rna # Bulk RNA-seq
python src/main.py --scrna # Single-cell RNA-seq
python src/main.py --protein # Protein structure
python src/main.py --epigenomics # ChIP-seq + ATAC-seq
python src/main.py --metagenomics # Metagenomics
python src/main.py --longread # Nanopore + PacBio
python src/main.py --assembly # Genome assembly
# Run ALL modules (full benchmark)
python src/main.py --all╔══════════════════════════════════════════╗
║ BioMark v0.5.0 🧬 ║
║ Bioinformatics Benchmark Tool ║
╚══════════════════════════════════════════╝ 🖥️ System Profile
Machine : MacBook Air M1
RAM : 8GB Unified Memory
SSD Free : 17GB
CPU : 8-core Apple M1 📊 BioMark Results
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
Module Score Capability
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
DNA/WES 15/100 🔴 RAM+SSD insufficient
RNA-seq 20/100 🔴 STAR needs 32GB RAM
scRNA-seq 10/100 🔴 Cell Ranger needs 32GB
Protein 15/100 🔴 AlphaFold needs 16GB
Epigenomics 10/100 🔴 Bowtie2 needs 16GB
Metagenomics 20/100 🔴 Kraken2 needs 16GB
Long Read 15/100 🔴 Minimap2 needs 16GB
Assembly 0/100 🔴 All assemblers fail
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
⭐ OVERALL 13/100
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━
An HTML report is generated automatically with interactive charts, per-step PASS/FAIL badges, hardware warnings, and time estimates.
| RAM | What You Can Run |
|---|---|
| 8GB | QC only (FastQC, MultiQC) |
| 16GB | WES single sample, RNA-seq with HISAT2 |
| 32GB | Full WES + RNA-seq with STAR |
| 36GB | WES + RNA-seq + scRNA-seq small datasets |
| 64GB | Full pipeline including medium scRNA-seq (30k cells) |
| 128GB+ | Large scRNA-seq (100k+ cells), genome assembly |
💡 Large databases (gnomAD ~1TB, Kraken2 ~100GB) can be stored on an external Thunderbolt SSD. BioMark detects this and gives specific recommendations automatically.
BioMark/
├── src/
│ ├── main.py # Main entry point + CLI
│ ├── system_info.py # Hardware detection
│ ├── report.py # HTML report generator
│ ├── utils.py # Utilities + JSON encoder
│ └── modules/
│ ├── DNA/ # WES pipeline benchmark
│ ├── RNA/ # Bulk RNA-seq + scRNA-seq
│ ├── Protein/ # AlphaFold2 + ESMFold + BLAST
│ ├── Epigenomics/ # ChIP-seq + ATAC-seq
│ ├── Metagenomics/ # Kraken2 + MetaPhlAn4 + QIIME2
│ ├── LongRead/ # Nanopore + PacBio pipelines
│ └── Assembly/ # SPAdes + Flye + Hifiasm
├── results/ # Benchmark results (JSON + HTML)
├── docs/
│ ├── CONTRIBUTING.md # Contribution guidelines
│ └── INSTALLATION.md # Detailed installation guide
├── requirements.txt
└── README.md
Run BioMark on your machine and contribute to our community database!
python src/main.py --all --shareThis generates a clean biomark_share_TIMESTAMP.json file with your hardware specs and scores.
To submit:
- Open a GitHub Issue: Submit My Score
- Title:
Score Submission — [Your Machine] [RAM]GB- Example:
Score Submission — MacBook Pro M5 Pro 36GB
- Example:
- Attach your
biomark_share_*.jsonfile
Your data helps researchers choose the right hardware for bioinformatics! 🧬
We welcome contributions from the bioinformatics community!
- Fork the repository
- Create a branch for your feature:
git checkout -b feature/new-module - Make your changes following the existing module structure
- Test your changes:
python src/main.py --your_module - Submit a Pull Request with a clear description
- 🆕 New pipeline modules (spatial transcriptomics, variant phasing)
- 🐛 Bug reports — please open a GitHub Issue
- 📊 Benchmark results from different machines (help us build the database!)
- 🌍 Testing on different operating systems and hardware
- Follow the existing module structure in
src/modules/ - Each step must return a
statusfield with✅ PASSor🔴 FAIL - Include honest hardware checks with RAM/SSD thresholds
- Add your module to
src/main.pyandsrc/report.py
See CONTRIBUTING.md for detailed guidelines.
BioMark does not yet have a peer-reviewed publication. If you use BioMark in your work, please cite it as: Shahbazi, A. BioMark: An open-source bioinformatics benchmark tool
for hardware evaluation (v0.5.0). GitHub (2026).
https://github.com/shahbazigenomics/BioMark
📌 DOI: 10.5281/zenodo.20738408 A tool paper is planned for submission to Bioinformatics or GigaScience.
Amir Shahbazi PhD Candidate in Genomics Specialization: Whole Exome Sequencing & Rare Variant Analysis in Familial Cases
This project is licensed under the MIT License. See LICENSE for details.
⭐ If BioMark helps your research or hardware decisions, please give it a star! It helps other researchers and students discover this tool.
BioMark simulates bioinformatics workloads for hardware benchmarking. It does not replace actual pipeline validation on real datasets.