GenomicBreedingModels

Genomic prediction and population genetics models (including multi-generation breeding simulations) for GenomicBreeding.jl.

Installation

This uses the R package BGLR. We therefore need to install R and the BGLR package first. To help with this, Conda maybe used by loading the environment file: misc/GenomicBreeding_conda.yml.

Then install GenomicBreedingModels.jl:

using Pkg
Pkg.add("https://github.com/GenomicBreeding/GenomicBreedingModels.jl")

If you wish to use neural network models, please install CUDA.jl on a GPU node. The install LuxCUDA.jl for an NVIDIA GPU or AMDGPU.jl for a node with an AMD GPU:

Log into a GPU node, e.g.:

sinteractive --job-name="CUDA_install" --account="account_name" --partition="gpu" --gres=gpu:1

Install CUDA.jl:

using Pkg
Pkg.add("CUDA")
using CUDA
CUDA.set_runtime_version!(v"12.8") # modify to match you CUDA version: see shell> nvidia-smi

Then restart Julia to download the CUDA_runtime:

using CUDA
if CUDA.functional(true)
    Pkg.add("LuxCUDA")
    using LuxCUDA
else
    Pkg.add("AMDGPU")
    using AMDGPU
end

Dev stuff:

REPL prelude

julia --threads 3,1 --load test/interactive_prelude.jl

Format and test

time julia --threads 3,1 test/cli_tester.jl