R package implementing Bayesian Weighted Quantile Sum Regression (BWQS) models and their extensions. The current version of the package contains three core functions:

1. bwqs(): BWQS for evaluating the association between exposure mixtures and a single outcome of interest which can be continuous (Gaussian), binary (Bernoulli), or count (Poisson).
2. dbwqs(): Dirichlet BWQS for evaluating the association between exposure mixtures and a compositional (multivariate) outcome consisting of proportions summing to 1.
3. hbwqs(): Hiearchical BWQS for evaluating the association between exposure mixtures and a single outcome of interest across multiple cohorts specified by the user. The same outcome types as bwqs() are supported.

Install the most recent version of xbwqs from GitHub via the remotes package: library(remotes) remotes::install_github('hasdk/xbwqs')

Please cite the references below when using the xbwqs package:

• Saddiki H, Warren JL, Lesseur C, Colicino E. “Compositional outcomes and environmental mixtures: the Dirichlet Bayesian Weighted Quantile Sum Regression” (2025). arXiv preprint; doi.org/10.48550/arXiv.2503.21428.
• Colicino E, Ascari R, Saddiki H, Mercedes-Nieves F, Pedretti NF, Huddleston K, Wright RO, Wright RJ. "Cross-cohort mixture analysis: a data integration approach with applciations on gestational age and DNA-methylation derived gestational age" (2024). Biometrical Journal; doi: 10.1002/bimj.202300270. PMID: 39473136.
• Colicino E, Pedretti NF, Busgang SA, Gennings C. "Per- and poly-fluoroalkyl substances and bone mineral density: Results from the Bayesian weighted quantile sum regression" (2020). Environmental Epidemiology; doi: 10.1097/EE9.0000000000000092. PMID: 32613152; PMCID: PMC7289141.