🧬 LycianWay

Automated bacterial genomic analysis · AI clinical interpretation · PCR primer design

Upload a single FASTQ file. Get a full clinical genomic report in ~15–45 minutes.

🌐 Landing page: https://cinnetcrash.github.io/GenomePlatform

---

What it does

1. Auto-detects MinION (long reads) vs Illumina (short reads)
2. QC — NanoPlot or FastP
3. Assembly — Flye (MinION) or Shovill/SPAdes (Illumina)
4. MLST typing — 100+ schemes via CGE MLST
5. AMR profiling — AMRFinderPlus
6. Annotation — Bakta
7. AI interpretation — Claude synthesises findings into a clinical narrative
8. PCR primer design — MAFFT conserved regions → Primer3
9. HTML report — downloadable, self-contained

All uploaded data is automatically deleted after 24 hours.

---

Quick start

Option A — Docker (recommended)

git clone https://github.com/cinnetcrash/GenomePlatform
cd GenomePlatform
cp .env.example .env          # add your ANTHROPIC_API_KEY
docker compose build
docker compose up -d
open http://localhost:8000

Option B — Direct (requires conda environments)

git clone https://github.com/cinnetcrash/GenomePlatform
cd GenomePlatform
export ANTHROPIC_API_KEY='sk-ant-...'
bash start.sh

---

System requirements

	Minimum	Recommended
CPU	8 cores	16+ cores
RAM	16 GB	32 GB
Disk	50 GB	500 GB SSD

Required tools (auto-installed via Docker): fastp · NanoPlot · Flye · Shovill · mlst · AMRFinderPlus · Bakta · MAFFT · Primer3

---

Project structure

GenomePlatform/
├── backend/
│   ├── main.py             # FastAPI app, all routes
│   ├── config.py           # Central configuration
│   ├── security.py         # File validation, path safety
│   ├── database.py         # SQLite job tracking
│   ├── cleanup.py          # 24h auto-deletion
│   ├── pipeline.py         # Read detection + analysis stages
│   ├── ai_interpreter.py   # Claude API integration
│   ├── primer_designer.py  # MAFFT + Primer3
│   └── report_generator.py # HTML report
├── frontend/
│   └── templates/index.html
├── docker/
│   ├── Dockerfile
│   └── environment.yml
├── docs/                   # GitHub Pages landing page
├── docker-compose.yml
├── start.sh
└── requirements.txt

---

Security

• File magic byte validation (no fake FASTQ.gz)
• Path traversal protection
• Rate limiting (5 uploads/min per IP)
• Max 3 concurrent jobs per IP
• subprocess always uses list args (shell=False)
• IP addresses stored as SHA-256 hashes (GDPR)
• All data auto-deleted after 24 hours

---

License

MIT © Gültekin Ünal — gultekinnunal@gmail.com

⚠️ Research use only. Not validated for clinical diagnostics.