Vertical Integration of Pharmacogenetics in Population PK/PD Modeling: A Novel Information Theoretic Method

Abstract

To critically evaluate an information-theoretic method for identifying gene–environmental interactions (GEI) associated with pharmacokinetic (PK), pharmacodynamic (PD), and clinical outcomes from genome-wide pharmacogenetic data. Our approach, which is built on the K-way interaction information (KWII) metric, was challenged with simulated data and clinical PK/PD data sets from the International Warfarin Pharmacogenetics Consortium (IWPC) and a gemcitabine clinical trial. The KWII efficiently identified both novel and known interactions for warfarin and gemcitabine. Interactions between herbal supplementation and VKORC1 genotype were associated with warfarin response. For gemcitabine-associated neutropenia, combination treatment with carboplatin and cytidine deaminase (CDA) 208G→A genotypes were identified as risk factors. Gemcitabine disposition was associated with drug metabolism–transporter interactions between deoxycytidine kinase (DCK) and the equilibrative nucleoside transporter (ENT). This novel approach is effective for detecting GEI involved in drug exposure and response and could enable integration of genome-wide pharmacogenetic data into the population PK/PD analysis paradigm.

Comparison of Nonlinear Mixed Effects Models and Noncompartmental Approaches in Detecting Pharmacogenetic Covariates

“…In the present work, we did not consider gene-gene and gene-environment interactions. Model-based approaches have been proposed in such contexts, and evaluated on real pharmacogenetic data sets (34). These methods or clusteringbased algorithms should be compared to penalised regression methods in simulations close to those presented in this work.…”

Machine learning in pharmacometrics: Opportunities and challenges

“…146 10 Previous work in our group demonstrated a novel approach to integrate pharmacogenomics data in PK/PD modelling using information theoretic approaches. 148 This method was used to simultaneously evaluate gene-environmental interactions using PK/PD, clinical outcomes and genome-wide pharmacogenetic data. Novel and known interactions between warfarin and gemcitabine were identified using the K-way interaction information metric.…”

“…In population PK/PD modelling, the covariate modelling component be an attractive target. For example, our work suggests that ML algorithms could be useful for building covariate models from high dimensional genotyping methods 64,148 . Similarly, large population‐based big data sets could be used extrapolate covariate models built from small PMX studies to minority populations and also to populations that are more diverse.…”

“…For example, our work suggests that ML algorithms could be useful for building covariate models from high dimensional genotyping methods. 64,148 Similarly, large population-based big data sets could be used extrapolate covariate models built from small PMX studies to minority populations and also to populations that are more diverse. The FDA and the PMX modelling community have recognized that the characteristics of the patients enrolled in drug development clinical trials are quite frequently not representative of the population at large.…”

Artificial Intelligence and Mechanistic Modeling for Clinical Decision Making in Oncology

“…For instance, instead of relying almost exclusively on linear covariate models (up to a transformation), highly nonlinear ML algorithms could be used to relate covariates and parameters of the structural model. 143,144 Another valuable approach is to derive quantitative metrics from simulation outputs of mechanistic modeling and use them as ML inputs for predictive purposes. 145 We propose to name such hybrid approaches combining big data and ML with mechanistic modeling "mechanistic learning" 85 (Figure 4).…”