Wen Rui Tan

About

Wen Rui Tan

I'm a PhD candidate in pharmacology working on population PK/PD modeling and model-informed precision dosing for children, and on what untargeted metabolomics and wearable sensor data add beyond drug concentrations.

Cincinnati, OH, United States · wenruiuc@gmail.com · LinkedIn · GitHub · Google Scholar

Education

  • PhD, Pharmacology, Physiology and Neurobiology

    University of Cincinnati College of Medicine, Cincinnati, OH

    Expected May 2027

    Dissertation: Advancing model-informed precision dosing via mechanistic and data-driven algorithms — an infliximab case study. Population PK/PD, ensemble machine learning, neural ODEs and deep reinforcement learning applied to dosing in pediatric Crohn's disease. Advisor: Tomoyuki Mizuno, PhD.

  • Bachelor of Pharmacy (Hons)

    Universiti Sains Malaysia, Penang, Malaysia

    Awarded September 2020

Experience

  • Graduate Research Assistant, Pharmacometrics

    Cincinnati Children's Hospital Medical Center, Cincinnati, OH

    August 2022 – present

    Population PK/PD modeling
    A pediatric population PK/PD model in NONMEM linking infliximab exposure to fecal calprotectin through an indirect response model, and population PK models for piperacillin to identify suitable doses for surgical populations. Both focus on the covariates that drive exposure and what they mean for dosing.
    Machine learning
    Ensemble models (XGBoost, random forest, CatBoost) that predict long-term biochemical remission from induction-phase data and individual PK/PD parameters, interpreted with SHAP, and a deep Q-network that treats dosing as a Markov decision process. All of it is benchmarked against the pharmacometric models.
    Tool development
    Interactive dashboards and standalone HTML reports in R and Python showing PK/PD simulation and modeling results (concentration-time profiles, diagnostics, target attainment) in a form clinical and research teams can read without a pharmacometrician in the room.
    Early-onset IBD
    Validating an existing population PK model in an early-onset IBD cohort, with a dashboard of descriptive statistics, diagnostic plots and individual patient PK reports.
    Clinical consultation
    Assistant PK consultant to clinical teams, analyzing therapeutic drug monitoring data to recommend dose adjustments for immunosuppressants (mycophenolic acid, tacrolimus, sirolimus) in high-risk pediatric patients.
    Study design
    A D-optimal design study in PopED to choose dupilumab sampling timepoints for pediatric eosinophilic esophagitis trials, maximizing information gain while minimizing patient burden.

  • Data Scientist Intern

    Enveda Therapeutics, Boulder, CO

    July 2026 – September 2026

    Linked untargeted metabolomics to wearable sleep, cardiovascular and activity data in healthy volunteers, using linear mixed-effects models with FDR control to find metabolite–wearable associations and characterize diurnal variability. Built the Python pipeline and dashboard. Also used metabolomics to classify ulcerative colitis vs Crohn's disease, and active disease vs remission, in a cohort already profiled by genomics, transcriptomics and proteomics.

  • Research Assistant

    University of Malaya, Kuala Lumpur, Malaysia

    March 2022 – May 2022

    Data cleaning and exploratory analysis of large patient datasets for a model-informed precision dosing initiative for vancomycin.

  • Research Assistant

    Universiti Sains Malaysia, Penang, Malaysia

    May 2021 – May 2022

    Systematic review and meta-analysis of isoniazid population PK models, establishing a framework for tuberculosis dosing across populations.

Skills

Pharmacometrics

  • NONMEM
  • mrgsolve
  • PopED
  • PsN
  • Pirana

Population PK/PD, exposure–response, indirect response models, Bayesian forecasting, optimal design, and dosing simulation for target attainment.

Machine learning

  • Gradient boosting
  • Random forest
  • CatBoost
  • SHAP
  • Neural ODEs
  • Variational autoencoders
  • Deep Q-networks

Used in my dissertation for remission prediction and dosing policy, always benchmarked against the mechanistic model.

Programming

  • R
  • Python
  • C++
  • Git

Shiny dashboards, reproducible analysis pipelines and standalone HTML reports. Examples are in the portfolio.

Therapeutic areas

  • Pediatrics
  • Infectious disease
  • Immunology
  • Rare disease

Biologics and antibiotics in critically ill and chronically treated children, and therapeutic drug monitoring.

Research interests

Population PK/PD

How drug exposure and response vary across a population, and what explains the variability.

Model-informed precision dosing

Turning those models into dose recommendations for individual patients, especially children, whose doses are too often extrapolated from adults.

Modeling and simulation

Indirect response models, exposure–response analysis and Monte Carlo simulation in NONMEM, R, mrgsolve and Python.

Clinical trial design

Optimal design methods for picking sampling schemes that get the most information with the least burden on patients.

Omics in clinical pharmacology

Untargeted metabolomics as a readout of disease state and treatment response, and the processing (filtering, non-detects, batch structure, repeated visits) a feature table needs before its associations are worth acting on.

Wearable and sensor data

Turning continuous sleep, cardiovascular and activity streams into exposures that can be joined to sparse biological samples, and separating within-person signal from diurnal patterns and non-wear.

Machine learning in pharmacometrics

Testing when ML actually does better than a mechanistic model, and when it only appears to.

Get in touch

I'm happy to hear about research collaborations in PK/PD modeling, methods questions in pharmacometrics, and internship or job opportunities.

Email LinkedIn GitHub Google Scholar Portfolio

Last updated September 2026