Serge Guzy, PhD

GuzySTATISTICAL EPIDEMIOLOGY, PHARMACOMETRICS,
PHARMACOKINETICS, & PHARMACODYNAMICS

PERSONAL PROFILE

  • 27 years in pharmacometrics, global modeling, and simulation for private companies and in academic settings.

EXPERTISE
Pharmacometrics; Statistical Epidemiology; Mathematical and Statistical Modeling; Clinical Trial Simulation; Pharmacokinetics; Pharmacodynamics; Compartment Analysis; Experiment Design; Algorithm Optimization; Non-continuous Likelihood-based Response Analysis (continuous/discrete Markov processes, generalized time/event analysis [single/multiple, and right-/left-/interval-centered events], count data analysis [Poisson and negative binomial], categorical multi-level analysis); PK/PD/Clinical Endpoint Simulation; Delay Differential Equations; Transit Compartment Expansion; Quantitative Risk Analysis; Portfolio Analysis Optimization; Management Strategy Development; Real Option Analysis; Stochastic NPV Analysis; Business Software Development; Conjoint Analysis; Product Design Optimization; Price Analysis; Product and Service Bundling; Brand Strategy Optimization; and Product Line Planning and Improvement. Programming expertise in Fortran 90, WinNonlin, Trial Simulator, S-ADAPT, NONMEM, PDx-MC-PEM, Phoenix NLME, and Monolix. 
Has made 18 presentations focusing on the forecaster (became the trial simulator) and 41 presentations focusing on the population approach, and has 29 publications.
Postdoctorate in Biostatistics and Pharmacokinetics from University of California-San Francisco, PhD in Biomedical Engineering from Technion – Israel Institute of Technology, and MS in Biomathematics and Chemistry from Weizmann Institute of Science – Rehovot, Israel. Also has a Chemical Physics License and Chemical Engineering degree from Brussels Engineering School.
PROFESSIONAL EXPERIENCE
President/CEO, POP_PHARM
Co-developer of the MC-PEM algorithm for Mixed Effect Modeling, which uses the expectation maximization algorithm combined with importance sampling, and Maximum a posteriori estimation. It is now the gold standard for Population Pharmacokinetic analysis and is proven to be unbiased, precise, stable, and robust. Consults in biostatistics, mathematical and statistical modeling, clinical trial simulation, pharmacokinetics, pharmacodynamics, compartment analysis, experiment design, and development of optimization algorithms. Clients include XOMA, Teva Israel, Obecure, NeuroDerm, Roche, Abbott, Omrix, Pharmapolaris, Hebrew University of Jerusalem (Yissum), Indiana University-USA, Azienda University-Italy, Genentech, Optimata, Regulus, Pharsight, Modigene Tech, Teva America, Medigen, Medistat, Bioforum, Yissum, ImmunoPharma, and Sunovion Pharmaceuticals.
Principal Scientist, Global Modeling and Simulation, XOMA Corporation
Provided PK/PD modeling expertise including clinical trial design, population PK/PD, statistical modeling, multiple imputation techniques, random number generator algorithm, power simulation, survival analysis simulation, Bayesian algorithms, randomization test algorithms, disease progression models, sequential trial designs, tumor growth modeling, assay statistical validation, and mathematical modeling of biological systems. Also contributed business modeling expertise including quantitative risk analysis; management strategies to reduce or ameliorate risk; portfolio analysis optimization using Monte Carlo simulation, decision theory, and real option analysis; Stochastic NPV analysis using both sensitivity and Monte Carlo analysis; business software development to improve quantitative project management; conjoint analysis; optimizing product design; analyzing price sensitivity; bundling product and service features; optimizing brand strategy; and improving product line planning.
Consultant for Winnonmix and Trial Simulator Testing, as well as Principal Scientist (SCI), Pharsight
Conducted classical parametric data analysis (sample size calculation, power, confidence intervals, hypothesis testing, distribution testing, Chi-square tests, Smirnov tests), non-parametric Bayesian analysis, non-parametric data analysis, non-linear mixed effect models (combined inter-subject/intra-subject variability), mathematical/statistical algorithm implementation (Matrix approach, non-linear conditional approach, and homoscedastic and heteroscedastic models), multivariate data analysis (continuous and discrete), multivariate parametric simulation, multivariate non-parametric simulation (flexible joint and marginal distributions, i.e., normal, lognormal , beta, Poisson, Weibull, multivariate normal, lognormal, and multivariate beta distributions), survival analysis, multiple events algorithms (survival analysis generalization), ANOVA, ANCOVA, bioequivalence, experiment design, and clinical trial simulation. Provided new random number generator techniques, computer algorithms, compliance and adherence models, Monte Carlo simulations, bootstrap techniques, multiple imputation techniques, and missing data algorithms. Developed strategies for data below quantifiable limits (BQL). Conducted compartment and non-compartment analysis (AUC, Cmax, tmax, etc.), physiological / pharmacokinetics modeling (discrete effect, link-effect, and indirect response, Emax, and inhibition models), logistic regression, polychotomous discrete data analysis (logistic regression generalization), pharmacokinetics / pharmacodynamics population statistical modeling, dosage regimen optimization, and Markov processes. Wrote testing checklists for accuracy and product performance for PK/PD software (trial simulator) that will soon be on the market. Designed and wrote code for Forecaster clinical trial software.
Assistant Professor, Biostatistics and Pharmacokinetics, University of California, San Francisco
Professor Affiliate, Pharmacometrics, University of Maryland
Adjunct Professor, Pharmacometrics, University of Minnesota
Adjunct Professor, Pharmacometrics, University of Colorado-Denver
Adjunct Professor, Pharmacometrics, University of Florida
 

Subscribe and Follow

Have the latest akta Pharmaceutical Development news delivered right to your inbox. 
And we will never share your information.