ncbi-tools

Docker image packaging NCBI bioinformatics tools for retrieving sequence data, metadata, and converting between formats. Designed for use as a runtime container in WDL/Cromwell genomics pipelines on platforms like Terra.

Included Tools

• entrez-direct - NCBI E-utilities command-line tools (esearch, efetch, elink, etc.)
• sra-tools - SRA Toolkit (sam-dump, fasterq-dump, etc.)
• samtools - Tools for manipulating SAM/BAM files
• picard - BAM/SAM manipulation and analysis
• biopython - Python library for biological computation
• perl-xml-simple - Perl XML parsing (used by entrez-direct)
• broad-tsv-converter - Asymmetrik TSV converter for NCBI data

Bundled Scripts

Script	Description
sra_to_ubam.sh	Fetch SRA reads and convert to unmapped BAM with full read group metadata
biosample-fetch_attributes.py	Fetch and harmonize BioSample metadata from NCBI (JSON + TSV output)
fetch_fastas_by_taxid_seqlen.sh	Retrieve FASTA sequences from NCBI by taxonomy ID and sequence length

Usage

Pull from GitHub Container Registry

docker pull ghcr.io/broadinstitute/ncbi-tools

Run interactively

docker run --rm -it ghcr.io/broadinstitute/ncbi-tools bash

Fetch BioSample metadata

docker run --rm -v $(pwd):/data ghcr.io/broadinstitute/ncbi-tools \
    python /opt/docker/scripts/biosample-fetch_attributes.py \
    --email your@email.org SAMN12345678 /data/output

Convert SRA to unmapped BAM

docker run --rm -v $(pwd):/data ghcr.io/broadinstitute/ncbi-tools \
    /opt/docker/scripts/sra_to_ubam.sh SRR1234567 /data/output.bam

Use in WDL pipelines

See WDL/task-sra_to_ubam.wdl for a ready-to-use WDL task definition.

Multi-Architecture Support

The image is built for both linux/amd64 (Intel/AMD) and linux/arm64 (Apple Silicon, ARM servers). Docker will automatically pull the correct architecture for your platform.

Modifying the Environment

Edit env.yaml to add or remove conda packages, then rebuild:

docker build -t ncbi-tools .