Skip to content

Latest commit

 

History

History
110 lines (94 loc) · 3.09 KB

README.md

File metadata and controls

110 lines (94 loc) · 3.09 KB
Raw

GWAS Protocol (as a container)

Usage

The container comes with all the software needed to generate summary statistics.

  1. You need to download the container file using one of the following commands. This will use approximately one gigabyte of storage.

    Container platform Version Command
    Singularity 3.x wget http://download.gwas.science/singularity/gwas-protocol-latest.sif
    Docker docker pull gwas.science/gwas-protocol:latest

  2. Next, start an interactive shell inside the container using one of the following commands.

    Container platform Command
    Singularity singularity shell --hostname localhost --bind ${working_directory}:/data --bind /tmp gwas-protocol-latest.sif
    Docker docker run --interactive --tty --volume ${working_directory}:/data --bind /tmp gwas.science/gwas-protocol /bin/bash

    3. Development

    To create a local development environment install Miniforge and create a .condarc file in your home directory with the following contents:

    channels:
  - conda-forge
  - bioconda

    Then update your .bashrc or .zshrc with mamba init. This will allow you to use the conda command.

    Next, install mamba using conda install mamba and then create the environment using the following command:

    micromamba create --name "gwas-protocol" \
  "conda-build" \
  "bcftools" "plink" "plink2" "tabix" "gcta" \
  "parallel" \
  "jupyterlab" "ipywidgets" \
  "python=3.12" "more-itertools" "psutil" "tqdm" "pyyaml" \
  "python-blosc2" "pyarrow" \
  "numpy" "scipy" "pandas" "threadpoolctl" "universal_pathlib" \
  "matplotlib" "seaborn" \
  "jax" "jaxlib=*=cpu*" "jaxtyping" "chex" "etils" "python-flatbuffers" \
  "mkl-include" "mkl" "c-blosc2" \
  "mypy" "pandas-stubs" "types-psutil" "types-pyyaml" "types-seaborn" "types-setuptools" "types-tqdm" \
  "pytest-benchmark" "pytest-cov" \
  "cython" "gxx_linux-64>=13" "gcc_linux-64>=13" "sysroot_linux-64>=2.17" "zlib" "gdb"

    Finally, install the gwas package using the following command:

    pip install --no-deps --editable "src/gwas"

    Benchmark

    data_path=/sc-projects/sc-proj-cc15-mb-enigma/genetics/development/opensnp
for sample_size in 100 500 3421; do
  mkdir -p "${sample_size}"
  pushd "${sample_size}" || exit 1
  benchmark --vcf $(for chromosome in $(seq 1 22); do echo ${data_path}/${sample_size}/chr${chromosome}.dose.vcf.zst; done) --output-directory . --method ramp --causal-variant-count 100 --simulation-count 1000 --seed 1000 --missing-value-pattern-count 10
  popd || exit 1
done