Sign inSign up
cdc-nanopore-offline logo

lsbnb/cdc-nanopore-offline

By lsbnb

•Updated 2 days ago

The discovery of mysterious pathogens

Image
0

74

lsbnb/cdc-nanopore-offline repository overview

⁠ClinMeta — Offline Clinical mNGS Pathogen Analysis

ClinMeta is a 100% offline-capable Nanopore mNGS pathogen analysis platform with a built-in Web GUI and multi-sample FIFO queue.


⁠Pull & Run (Quick Start)

⁠1. Pull the Image & Create ./data/
docker pull <your-dockerhub-user>/clinmeta:latest
mkdir -p data/{input,output,databases,references}
⁠2. Start the Container

Option A — Standard Offline Run (mount existing local databases under ./data/):

docker run -d \
  --name clinmeta-offline \
  -p 8080:8080 \
  -v "$PWD/data/databases:/data/databases" \
  -v "$PWD/data/references:/data/references" \
  -v "$PWD/data/input:/data/input:ro" \
  -v "$PWD/data/output:/data/output" \
  <your-dockerhub-user>/clinmeta:latest

Option B — First-Time Auto-Download (downloads reference bundle from Zenodo + Kraken2 PlusPF from AWS S3 on startup):

docker run -d \
  --name clinmeta-offline \
  -p 8080:8080 \
  -e CDC_ZENODO_BUNDLE_URL="<your-zenodo-bundle-url>" \
  -e CDC_AUTO_DOWNLOAD_K2=true \
  -v "$PWD/data/databases:/data/databases" \
  -v "$PWD/data/references:/data/references" \
  -v "$PWD/data/input:/data/input:ro" \
  -v "$PWD/data/output:/data/output" \
  <your-dockerhub-user>/clinmeta:latest

Open http://<server-ip>:8080 in your browser to queue samples and view results.


⁠Directory Layout (./data/)

./data/
  ├── input/       # <sample_id>/dna/*.fastq.gz & <sample_id>/cdna/*.fastq.gz
  ├── output/      # Analysis results, BAMs, HTML/Sankey reports, jobs.db
  ├── databases/   # kraken2/k2_pluspf_20261022/ (~117 GB) + viral_refseq_prot/ (~331 MB)
  └── references/  # human-masked* (~20 GB) + pathogen_panel/ (~136 MB)

⁠Pipeline & Included Tools

  • Workflow: chopper QC → hostile host removal → Kraken2 → competitive minimap2 validation → DIAMOND blastx unknown-virus screen → Sankey & HTML reports
  • Tools: Python 3.12 · chopper 0.15.0 · hostile 2.0.2 · minimap2 2.31 · kraken2 2.17.2 · samtools 1.25 · diamond 2.1.9 · seqkit 2.14 · pigz 2.8

⁠Environment Variables & Requirements

VariableDefaultPurpose
CDC_PORT8080Web GUI port
CDC_INPUT_DIR / CDC_OUTPUT_DIR/data/input / /data/outputSample I/O paths
CDC_DB_DIR / CDC_REF_DIR/data/databases / /data/referencesDatabase & reference paths
CDC_ZENODO_BUNDLE_URL""Auto-download ref bundle if missing
CDC_AUTO_DOWNLOAD_K2falseAuto-download Kraken2 PlusPF from AWS S3
CDC_BUILD_THREADS16Threads for startup index building
  • Hardware: ≥ 128 GB RAM (for Kraken2 PlusPF), ≥ 200 GB SSD/NVMe, 16+ CPU cores

Tag summary

Content type

Image

Digest

sha256:78a9a4381…

Size

891.5 MB

Last updated

2 days ago

docker pull lsbnb/cdc-nanopore-offline