GenomePlatform automates bacterial genome assembly, AMR profiling, MLST typing, and AI interpretation — all from a single FASTQ file. No coding required.
From raw reads to actionable clinical insights — automatically.
Automatically distinguishes MinION long reads from Illumina short reads based on read length distribution. No configuration needed.
Flye for long reads, Shovill/SPAdes for Illumina. Produces high-quality assemblies ready for downstream analysis.
AMRFinderPlus detects resistance genes, with class, subclass, and identity scores. Carbapenemases, ESBL, colistin — all covered.
Sequence type determination for 100+ schemes. Identifies high-risk clones like ST131, ST258, ST11 instantly.
Claude AI synthesizes all findings into a clinical narrative: species prediction, risk level, treatment implications, and epidemiological context.
9 automated stages, ~15–45 minutes end-to-end
All uploaded files and analysis results are automatically deleted after 24 hours.
Run on your own hardware. Your genomic data never leaves your institution.
IP addresses are hashed and never stored in plain text. No analytics, no cookies.
git clone https://github.com/cinnetcrash/GenomePlatform
cd GenomePlatform
cp .env.example .env
# Edit .env and add your ANTHROPIC_API_KEYdocker compose build
docker compose up -d
# Platform running at http://localhost:8000Or without Docker (requires conda environments):
export ANTHROPIC_API_KEY='sk-ant-...'
bash start.sh
Open http://localhost:8000,
drag and drop your FASTQ file, and the pipeline starts automatically.
You'll receive a detailed HTML report with QC metrics, AMR profile,
and AI interpretation.