Quantitative Systems Pharmacology
Mechanistic models, virtual populations and digital twins for translational research and model-informed drug development
The Quantitative Systems Pharmacology (QSP) Group at COSBI develops mechanistic computational models to understand complex disease biology, predict treatment response and support the development of new therapeutic strategies.
We combine systems biology, pharmacokinetics and pharmacodynamics, PBPK modelling, mathematical modelling and data-driven methods to connect mechanisms across biological scales — from molecular and cellular processes to tissue exposure, disease progression and clinical outcomes.
Our work spans collaborations with pharmaceutical and biotechnology companies, non-profit research organizations, academic institutions and clinical centres, alongside methodological research aimed at advancing the way QSP models are developed, evaluated and used.
The group is led by Federico Reali, PhD, Group Leader in Quantitative Systems Pharmacology at COSBI.
1. Translational QSP and Model-Informed Drug Development
We develop quantitative models to support decisions throughout the drug development process, from the interpretation of preclinical evidence to clinical translation.
Depending on the scientific question, we integrate mechanistic disease models with PK/PD and physiologically based pharmacokinetic models (PBPK/mPBPK) to investigate drug exposure, target engagement and pharmacological response across species, tissues and patient populations.
Our work with pharmaceutical and biotechnology partners includes applications such as:
- translation from preclinical experiments to humans;
- prediction and optimization of clinically relevant dose ranges;
- integration of pharmacokinetics, pharmacodynamics and disease mechanisms;
- tissue- and site-of-action exposure modelling;
- assessment of treatment response and sources of variability;
- development of virtual patients and populations;
- quantitative comparison of alternative therapeutic strategies;
- support for model-informed drug development and regulatory-facing analyses.
The group works across different therapeutic modalities, including small molecules, biological therapeutics, antibodies, and novel drug-delivery approaches.
Our objective is not simply to reproduce available data, but to build models that can be used to ask new questions, test hypotheses and generate predictions that are relevant for drug development.
2. Mechanistic Disease Modelling and Collaborative Research
A substantial part of our research is developed together with academic, clinical and non-profit partners, where quantitative modelling can help connect biological knowledge with experimental and clinical observations.
Tuberculosis and global health
We develop translational pharmacology models for tuberculosis drug development, including minimal-PBPK approaches, lung and lesion pharmacokinetics, spatial models of drug penetration and multiscale representations of granuloma biology.
Our collaboration with the Gates Medical Research Institute has led to the development of stormTB, an openly accessible simulator for anti-tuberculosis pharmacokinetics, as well as subsequent work extending these approaches towards lesion-level drug exposure, target attainment and translational dose prediction.
These projects combine pharmacology with spatial and mechanistic modelling, including PBPK, partial differential equations and agent-based modelling, to investigate how drug distribution within pulmonary lesions may influence treatment efficacy.
Neurodegenerative diseases
Neurodegeneration is another major research area for the group.
Our work includes mechanistic modelling of alpha-synuclein aggregation in Parkinson’s disease, investigating how molecular processes and lipid environments contribute to protein aggregation and the formation of potentially toxic species.
We are also developing approaches for GBA1-associated Parkinson’s disease, integrating sphingolipid metabolism, biomarkers and clinical data with the longer-term objective of constructing patient-specific mechanistic models and digital twins for disease stratification and therapeutic investigation.
Rare and complex diseases
The group has extensive experience in QSP modelling of lysosomal storage disorders, including Gaucher disease and related sphingolipidoses.
These models provide a quantitative framework for studying metabolic alterations, disease heterogeneity and therapeutic intervention, and also represent an important methodological foundation for our current research on patient-specific models and mechanistic digital twins.
Across these disease areas, we work closely with clinicians, experimental scientists and disease experts so that model development remains anchored to relevant biological and clinical questions.
3. Virtual Populations, Digital Twins and Next-Generation QSP
Alongside disease- and drug-specific projects, we develop methods that address fundamental challenges in quantitative systems pharmacology.
Virtual populations
Biological variability is central to translating mechanistic models from an “average” system to heterogeneous patient populations.
We develop and benchmark computational approaches for generating plausible patients and virtual populations, including optimization, Bayesian sampling and surrogate-assisted methods.
Our research investigates how different strategies affect:
- representation of biological variability;
- parameter non-identifiability;
- computational efficiency;
- population calibration;
- uncertainty in model predictions;
- robustness and reproducibility of QSP analyses.
Rather than assuming that a single computational strategy is optimal for every problem, we study how model structure, available data and context of use should guide the construction of virtual populations.
Mechanistic digital twins
We are extending these concepts towards mechanistic digital twins: quantitative models conditioned on information from individual patients and designed to reproduce relevant aspects of their biological state.
Our research investigates how mechanistic knowledge, longitudinal clinical information, biomarkers and data-driven approaches can be integrated to support patient stratification and individualized simulations.
AI, model credibility and regulatory science
We are also exploring how machine learning and artificial intelligence can complement mechanistic modelling, for example through surrogate models, parameter-space exploration, knowledge extraction and hybrid mechanistic/data-driven approaches.
At the same time, increasing model complexity requires clear approaches for establishing model credibility, transparency and reproducibility.
The group therefore contributes to international discussions around QSP best practices, credibility assessment and regulatory use of mechanistic models, including activities within the International Society of Pharmacometrics (ISoP) and engagement with initiatives concerning the qualification and reporting of mechanistic models.
The group also contributes to the broader Italian life-sciences ecosystem through ALISEI – the Italian National Life Sciences Cluster, including activities related to digital health, health data, telemonitoring and the translation of computational innovation into healthcare practice.
Selected Research Areas
Quantitative Systems Pharmacology
Mechanistic integration of disease biology, pharmacology and clinical data.
PBPK and Translational Pharmacology
Cross-species translation, tissue exposure, target-site pharmacokinetics and dose selection.
Virtual Populations
Computational methods to represent biological and clinical variability in mechanistic models.
Mechanistic Digital Twins
Patient-conditioned models for stratification, individualized simulation and translational research.
Tuberculosis
Drug distribution, lung and lesion pharmacokinetics, spatial modelling and treatment optimization.
Neurodegeneration
Parkinson’s disease, GBA1-associated Parkinson’s disease, and alpha-synuclein biology.
Rare Diseases
Lysosomal storage disorders, Gaucher disease and sphingolipid metabolism.
Advanced Therapeutics
Biological agents, antibodies, CAR-T therapies and innovative delivery approaches.
Model Credibility and Regulatory Science
Context of use, model evaluation, reporting standards and trustworthy application of mechanistic models.
Group Leader
Federico Reali, PhD
Group Leader – Quantitative Systems Pharmacology
Federico Reali holds a PhD in Mathematics from the University of Trento and leads the Quantitative Systems Pharmacology group at COSBI.
His research focuses on the development and application of mechanistic mathematical models, QSP, virtual populations and digital twins to support drug development and translational biomedical research.
He coordinates multidisciplinary projects with pharmaceutical companies, non-profit organizations, academic groups and clinical collaborators across therapeutic areas including infectious diseases, neurodegeneration and rare diseases.
His current methodological interests include virtual population generation, patient-specific mechanistic modelling, integration of AI with QSP, and the assessment of model credibility for translational and regulatory applications.
He is also involved in university teaching at the University of Trento, with courses in statistics, mathematical modelling and systems biology, and supervises MSc and PhD research projects in quantitative and computational biomedicine.
Federico is also involved in international QSP initiatives within the International Society of Pharmacometrics and serves as Vice President of the Scientific Committee of ALISEI, the Italian National Life Sciences Cluster.
Collaborate with us
We are interested in collaborations where mechanistic modelling can help transform complex biological, preclinical or clinical data into quantitative hypotheses and actionable predictions.
We collaborate with pharmaceutical and biotechnology companies, academic and clinical research groups, non-profit organizations and international research initiatives on both applied drug-development projects and fundamental methodological research.
Interested in working with the COSBI QSP Group?
Contact us to discuss potential research collaborations, scientific projects, PhD opportunities or applications of quantitative systems pharmacology.