Skip to content

installation.md

Latest commit

 

History

History
143 lines (95 loc) · 3.58 KB

installation.md

File metadata and controls

143 lines (95 loc) · 3.58 KB

Installation

This guide covers installing ROBIN and its dependencies from source for this repository (ROBIN).

For running workflows (MinKNOW settings, BAM read limits, alignment, and CLI usage), see the repository README—especially BAM read limit and MinKNOW settings and Usage. This page focuses on installation.

ROBIN with Little John

Prerequisites

  • Git and Git LFS
  • Conda (Miniconda or Anaconda)
  • Python 3.12 (supplied by the robin conda environment)

Recommended:

  • 64 GB RAM or more for production runs
  • CPU/GPU per Oxford Nanopore guidance for your sequencer

Docker is optional (some third-party or containerized toolchains may use it).

Step 1: Clone the repository

Clone including submodules:

git clone --recursive https://github.com/LooseLab/ROBIN.git
cd ROBIN

If you already cloned without submodules:

git submodule update --init --recursive

Step 2: Create the conda environment

Use the environment file at the repository root:

File Purpose
robin.yml Main environment: Python 3.12, scientific stack, bioinformatics tools, R/Bioconductor (Linux and macOS).
robin_linux_extras.yml Linux only, optional: merge after creating the env if you hit libstdc++ / CXXABI_1.3.15 issues (see README Common issues). 
conda env create -f robin.yml
conda activate robin

Notes:

  • pyproject.toml requires Python >=3.12; this environment matches that.
  • Prefer a fresh environment for this codebase rather than reusing an older ROBIN env from past releases.

If the robin environment already exists

robin.yml sets name: robin. If conda env create -f robin.yml reports the environment already exists:

Option A — Update in place

conda env update -n robin -f robin.yml --prune
conda activate robin

Option B — Remove and recreate (clearest if the old env is stale or mixed)

conda deactivate
conda env remove -n robin
conda env create -f robin.yml
conda activate robin

Option C — New name

conda env create -f robin.yml -n robin_littlejohn
conda activate robin_littlejohn

On Linux, if you see CXXABI_1.3.15 / wrong libstdc++:

conda env update -n robin -f robin_linux_extras.yml

(See README Common issues for details.)

Step 3: Install ROBIN (editable)

From the repository root:

pip install -e .

This installs the robin CLI from your working tree.

Step 4: Download models and ClinVar

After installation, fetch bundled assets (SHA256-verified; set GITHUB_TOKEN if assets are on private GitHub):

robin utils update-models
robin utils update-clinvar

Private GitHub:

export GITHUB_TOKEN=your_personal_access_token
robin utils update-models

Force re-download models if needed:

robin utils update-models --overwrite

Step 5: Verify

robin --help
robin list-job-types

Troubleshooting

Issue What to do
Missing submodules git submodule update --init --recursive
Model / ClinVar download failures Set GITHUB_TOKEN if required; retry with robin utils update-models --overwrite; run robin utils update-clinvar
Wrong conda env conda env list and activate the env you created from robin.yml

Next steps