About Me
I’m an epidemiologist and statistician formerly working in chemical safety and chronic disease research at the Seattle Children’s Research Institute, and now looking for opportunities in pharmaceutical research and development.
My primary experience is in both developing and applying statistical methods to investigate the determinants of chronic disease. I have training in Bayesian and frequentist statistics, machine learning, epidemiology, toxicology, immunology, bioinformatics (transcriptomics, scRNA, metabolomics, etc.), endocrinology, and neurology.
Methods Development/Teaching
- R package wqspt: Using a novel permutation test method to improve the statistical properties of the weighted quantile sum regression, which determines the additive combined association of multiple exposures with a single outcome.
- R package ResidualKMeans: Method for unsupervised clustering of mixed type continuous and categorical data while controlling for structure induced by a grouping variable such as cohort in a multi-cohort study.
- R package impgrplasso: Combining methods involving a group LASSO penalty for multiply imputed data into one function to perform LASSO regression on data with missingness.
- R package DrewDayRFunctions: Functions with relevance for epidemiologic analyses including functions to performing many regressions, nonlinear mediation, nonlinear regressions, and exploratory data analysis.
- R teaching Introduction to R: Document to teach a student just starting with R to go from zero to developing useful functions and loops.
Research Interests
- Omics prediction of chronic disease: Using omics data to predict chronic disease patterns to discover shared mechanisms (Placental gene transcription and child cognitive and behavioral outcomes (abstract))
- Multi-outcome health profiles: Using phenotypes of disease outcome co-occurrence to indentify risks and mechanisms shared between seemingly unrelated diseases (Residual K-means paper) (see also my experiments with autoencoders at VAECM_Final)
- Mixture exposures: Advancing mixture exposure modelling to understand the combined health impact of common environmental factors (Weighted Quantile Sum regression with Permutation Test (WQSPT) methods paper)
- Environmental Epidemiology: exploring associations between chemical and other stressor exposures and later life chronic disease:
- Childhood melamine and kidney health outcomes
- Prenatal phthalate plasticizers and child behavioral disorders
- Prenatal maternal sex hormones and child behavioral disorders
- Ozone air pollution and cardiopulmonary health in adults
- Electrostatic precipitator air filtration and cardiopulmonary health in adults