Powering Research Through Innovative Methods for Mixtures in Epidemiology (PRIME)
Open-access software (primarily R code), related tools to enable broader application of the novel methods developed under PRIME, and additional work related to environmental mixtures methodology are shared. Questions on code can be addressed to PRIME investigators/collaborators listed on specific pages or to the PRIME Program Director, Bonnie R. Joubert.
GitHub
The NIEHS PRIME GitHub includes shared software in development, or previously developed and utilized, or expanding in a PRIME project. It also includes shared public and simulated datasets used for PRIME methods development.
Software
The following represent new software developed to address the statistical analysis of mixtures data. Package author/contact noted in parentheses.
- Bayesian Infinite Factor Modeling (Poworoznek), including the most efficient way to run models from Factor analysis for Interactions (see below)
- Bayesian partially Supervised Sparse and Smooth Factor Analysis (BS3FA) (Dudson, Moran)
- BMC (Jin, Dunson)
- BMIM (McGee)
- COLRNS (Chen)
- CVEK (Deng, Coull)
- DLMtree (Mork, Wilson)
- Factor analysis for Interactions (FIN) (Ferrari, Dunson)
- MixSelect (Ferrari, Dunson)
- MVNImpute (Chen)
- Nlinteraction (Antonelli)
- Perturbed factor model (Arkaprava, Dunson)
- Regimes (Wilson)
Data
- HHEAR data center
- NHANES dataset from 2019 Columbia Mixtures Workshop
- Simulated data from 2015 NIEHS Mixtures Workshop
to Top