Hub

Pangolin

SARS-CoV-2 lineage assignment using the Pango nomenclature

Pangolin assigns SARS-CoV-2 genome sequences to Pango lineages for genomic surveillance. Given one or more consensus genomes in a FASTA file, it aligns them, runs the assignment pipeline, and writes a lineage report.

Key features:

  • Assign SARS-CoV-2 sequences to Pango lineages
  • Bundled lineage and designation data (pdata)
  • CSV report with lineage, conflict, and QC status per sequence

Configuration

name: pangolin
tags:
    - genomics
steps:
    - name: job
      platform: docker
      mode: parallel
      image: ghcr.io/dxflow-ai/pangolin:latest
      volumes:
          - name: input
            host: ./input
            container: /data/input
            mode: ro
          - name: output
            host: ./output
            container: /data/output
      env:
          - INPUT=/data/input/sequences.fasta
          - THREADS=4
          - EXTRA=
      resources:
          cpu: "4"
          memory: 8G

[volume]
job.input = ./input
job.output = ./output

[env]
job.INPUT = /data/input/sequences.fasta
job.THREADS = 4
job.EXTRA =

[resource]
job.cpu = 4
job.memory = 8G

{
    "arch": ["amd64"],
    "image": "ghcr.io/dxflow-ai/pangolin:latest",
    "version": "4.3.1",
    "minimum": {
        "cpu": 2,
        "memory": "4G",
        "storage": "10G"
    }
}

Usage

1. Prepare data

Upload a FASTA file of one or more SARS-CoV-2 consensus genomes:

# Create input/output directories
mkdir -p input output

# Upload your sequences
dxflow artifact upload /local/sequences.fasta input/

2. Deploy

dxflow workflow create --identity pangolin pangolin.yml

3. Start (with optional tuning)

The step reads INPUT (the query FASTA), THREADS, and EXTRA (extra pangolin flags, e.g. --analysis-mode fast). Override them per run:

# Start with defaults
dxflow workflow start pangolin

# Or point at another file and pass extra flags
dxflow workflow start pangolin \
    --override env.job.INPUT=/data/input/genomes.fasta \
    --override env.job.EXTRA=--analysis-mode\ usher

4. Retrieve results

dxflow artifact download output/ /local/pangolin-results/

Output files

  • lineage_report.csv - one row per input sequence: taxon, lineage, conflict, scorpio_call, qc_status, and the tool/data versions used

Notes

  • Input should be near-complete SARS-CoV-2 genomes; short or low-coverage sequences are still reported but marked fail in qc_status.
  • The image bundles the lineage data (pdata), so assignment runs offline — no network download is needed at runtime.

References

SAMtools

Read, write, and manipulate SAM, BAM, and CRAM alignment files

Molecular

Pre-configured workflows for simulating molecular behavior and interactions