Regularized mixed-effects models in Julia

Overview

HighDimMixedModels.jl is a package for fitting regularized linear mixed-effect models on high-dimensional omics data. These models can be used to analyze hierarchical, high dimensional data, especially useful for situations in which the number of predictors exceeds the number of samples ($p > n$).

For fitting the model, the package implements the coordinate gradient descent algorithm found in

Schelldorfer, Jürg, Peter Bühlmann, and SARA VAN DE GEER. "Estimation for high‐dimensional linear mixed‐effects models using ℓ1‐penalization." Scandinavian Journal of Statistics 38.2 (2011): 197-214.

and

Ghosh, A., & Thoresen, M. (2018). Non-concave penalization in linear mixed-effect models and regularized selection of fixed effects. AStA Advances in Statistical Analysis, 102, 179-210.

Because of its superior estimation performance, the smoothly clipped absolute deviation (SCAD) is the default penalty, but the $\ell_1$ penalty (LASSO) proposed in the original source is also available.

Installation and usage

See the package documentation for details on how to install and use the package.

Errors and contributions

To report a bug or propose a new feature, please open an issue in the issue tracker.

We welcome contributions. User who are interested in contributing code are asked to follow the instructions found in CONTRIBUTING.md

Citation

If you use HighDimMixedModels.jl in your work, we kindly ask that you cite

@article{Gorstein2024,
author = {Gorstein, E. and Aghdam, R. and Sol'{i}s-Lemus, C.},
year = {2024},
title = {{HighDimMixedModels.jl: Robust High Dimensional Mixed Models across Omics Data}},
journal = {In preparation}
}