The DIGPHAT consortium, led by JB Woillard (INSERM, CHU Limoges) and C Battail (CEA Grenoble), aims to develop pharmacological digital twins—computational representations of patients combining mechanistic models (PK, PK/PD, PBPK) with AI-driven approaches using multi-source, multi-scale clinical and biological data. These digital twins are designed to predict individual treatment responses, optimize drug dosing, and improve the personalization of therapies.
The project is structured around the creation of meta-models that integrate mechanistic and data-driven methods across different time scales and clinical contexts. Three major public health challenges are addressed as case studies: solid organ transplantation, oncology (kidney and breast cancers), and infectious diseases (antibiotic resistance and fungal infections). These use cases were chosen for their diversity in treatment durations, toxicity mechanisms, and exposure-response dynamics.
A key innovation lies in the use of synthetic patient data generated by advanced machine learning techniques, which supports data sharing in compliance with GDPR. In addition, this approach may reduce the number of subjects required for randomized clinical trials (RCTs), especially by augmenting underrepresented subgroups.
DIGPHAT also explores the role of causal machine learning for estimating individualized treatment effects (CATE), enabling virtual trial simulations and improving decision-making in real-world clinical settings. While digital twins are not a replacement for RCTs, they offer complementary value: enabling faster patient selection, hypothesis generation, and early efficacy/toxicity prediction. The consortium involves multidisciplinary experts from pharmacology, integrative genomics, AI, and mathematical modeling, leveraging their synergy to push the boundaries of precision medicine.

This seminar will be of particular interest for researchers in biomedical/clinical science, data science and AI in health. We look forward to your participation and to an inspiring discussion.

​Contact : Christophe Battail​​​

​