The R Package BHAM: Fast and Scalable Bayesian Hierarchical Additive Model for High-dimensional Data

TL;DR

BHAM is an R package that offers fast, scalable Bayesian hierarchical additive modeling tools for high-dimensional clinical and genomic data, enabling flexible analysis and variable selection.

Contribution

This paper introduces BHAM, a new R package implementing scalable Bayesian additive models with spike-and-slab LASSO priors for high-dimensional data analysis.

Findings

Efficient algorithms for high-dimensional additive models.

Utility functions for model selection and visualization.

Application to molecular data for disease prediction.

Abstract

BHAM is a freely avaible R pakcage that implments Bayesian hierarchical additive models for high-dimensional clinical and genomic data. The package includes functions that generalized additive model, and Cox additive model with the spike-and-slab LASSO prior. These functions implement scalable and stable algorithms to estimate parameters. BHAM also provides utility functions to construct additive models in high dimensional settings, select optimal models, summarize bi-level variable selection results, and visualize nonlinear effects. The package can facilitate flexible modeling of large-scale molecular data, i.e. detecting susceptible variables and infering disease diagnostic and prognostic. In this article, we describe the models, algorithms and related features implemented in BHAM. The package is freely available via the public GitHub repository https://github.com/boyiguo1/BHAM.