APOD: reasoning-guided agentic population ordinary differential equation discovery for pharmacological digital twins
Abstract
Establishing ordinary differential equations (ODEs) describing population data is a fundamental part of mathematical modeling in pharmacology, crucial to developing digital twins. However, doing so from sparse, noisy data is a slow, expert-driven task. Existing automated methods either search a restricted model space or ignore population inter-individual variability. Here we introduce APOD (Agentic Population ODE Discovery), a language-model agent that iteratively reasons over biological knowledge and fit diagnostics in an open-ended search space to discover a population digital twin (PDT), i.e., a shared ODE system with between-subject variability. On synthetic pharmacokinetic and tumor-dynamics benchmarks, APOD recovered ground-truth structures in 94-100\% of runs, 12-fold faster in median than an established library-based search. On real cohorts it converged to valid structures, and proposed a PDT of radioligand-therapy-induced platelet dynamics that predicts thrombocytopenia from first-cycle data and simulates alternative dosing schedules that lower the predicted risk of toxicity.
Related
- Chaotic dynamics in blooms of Ostreopsis cf. ovata modelled from environmental time series using ordinary differential equations systems
- Differential Adaptation of Osteocyte Lacunar Morphology at Bone (Re)modeling Sites under Mechanical Loading and Pharmacological Therapies in a Mouse Model of Osteoporosis
- Multimodal reasoning for broadly neutralizing antibody discovery from label-free human B cell repertoires across virus families
- CircuitATLAS: Agentic reasoning over a systems neuroscience knowledge graph for target discovery in circuitopathies
- Discovery of a Virulent Population of Meloidogyne graminicola Overcoming the Rice Resistance Gene MG1 and Population Genetic Analysis Based on mtCOI Sequences
BibTeX
@misc{ferrara2026apod,
title = {{APOD: reasoning-guided agentic population ordinary differential equation discovery for pharmacological digital twins}},
author = {Romain Ferrara and Martin Soucail and Victor Gertner and Adil Moussali and Joris Cocquebert and Sandrine Oziel-Taieb and Julien Nicolas and Florence Gattacceca and Mihaela van der Schaar and Sébastien Benzekry},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.06227},
eprint = {2610.06227},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.06227},
}
RIS
TY - UNPB
TI - APOD: reasoning-guided agentic population ordinary differential equation discovery for pharmacological digital twins
AU - Romain Ferrara
AU - Martin Soucail
AU - Victor Gertner
AU - Adil Moussali
AU - Joris Cocquebert
AU - Sandrine Oziel-Taieb
AU - Julien Nicolas
AU - Florence Gattacceca
AU - Mihaela van der Schaar
AU - Sébastien Benzekry
PY - 2026
DA - 2026/10/05
PB - arXiv
DO - 10.48550/arxiv.2610.06227
UR - https://arxiv.org/abs/2610.06227
AB - Establishing ordinary differential equations (ODEs) describing population data is a fundamental part of mathematical modeling in pharmacology, crucial to developing digital twins. However, doing so from sparse, noisy data is a slow, expert-driven task. Existing automated methods either search a restricted model space or ignore population inter-individual variability. Here we introduce APOD (Agentic Population ODE Discovery), a language-model agent that iteratively reasons over biological knowledge and fit diagnostics in an open-ended search space to discover a population digital twin (PDT), i.e., a shared ODE system with between-subject variability. On synthetic pharmacokinetic and tumor-dynamics benchmarks, APOD recovered ground-truth structures in 94-100\% of runs, 12-fold faster in median than an established library-based search. On real cohorts it converged to valid structures, and proposed a PDT of radioligand-therapy-induced platelet dynamics that predicts thrombocytopenia from first-cycle data and simulates alternative dosing schedules that lower the predicted risk of toxicity.
ER -