Statistical analysis and modeling of vaccines
Slides: https://www.andreashandel.com/presentations/
2025-10-13
A few topics related to vaccine modeling and analysis
- Introduction
- Statistical analysis of the role of dose for norovirus antibody responses
- Mechanistic (QSP) modeling for vaccines
- Epidemiological vaccine modeling
- Tools
Introduction
Current approaches to dose choice
- A few doses are explored in phase I/II trials, one is chosen for phase III and licensure.
Example of BioNTech/Pfizer COVID Vaccine
Claim: Knowing in more detail how dose impacts host response following vaccination might help optimize vaccines.
Statistical analysis of the role of dose for norovirus antibody responses
Motivation
- Most vaccines aim to induce strong antibody responses.
- For a future norovirus vaccine, it is important to understand how dose impacts antibody responses.
- It is also informative to understand if there are differences in antibody kinetics following infection vs. vaccination.
Study design
Models
Exponential decay model: y1(t)=p11+e(−g1∗(t−k1))e−d1∗t
Power-law decay model: y2(t)=p21+e(−g2∗((t+1)−k2))(t+1)−d2
Vaccination antibody time-series
Hierarchical/mixed-effects modeling framework
Vaccination antibody time-series
Vaccination antibody time-series
Vaccination antibody early response
Vaccination antibody durability
Mechanistic (QSP) modeling for vaccines
Objectives
For this project, the goals were to:
- Develop a conceptual framework combining data with mechanistic models to investigate the impact of dose on infection dynamics.
- Link the models to vaccine outcomes (protection and morbidity).
- Illustrate how to use this framework to predict outcomes for a large range of doses.
We focused on viral infections as a stand-in for live attenuated vaccines.
Influenza (vaccine)
Modeling dose and immune response
Modeling dose and immune response
Uninfected cells˙U=−bUVInfected cells˙I=bUV−dIIDead cells˙D=dIIVirus˙V=pI1+sFF−(dVV+k′AA+b′U)VInnate response˙F=pF−dFF+gF(Fmax−F)VV+hVB cells˙B=FVFV+hFgBBAntibodies˙A=rAB−dAA−kAAV
Modeling dose and immune response
Mapping antibodies to protection
We want to know protection. We can map antibodies to it. We use this relation: P=1−1/(1+ek1(log(A)−k2))
Mapping innate response to morbidity
We want to know morbidity (side effects/symptoms). We can map the innate response to it. M=∫aFcb+Fc
Dose for influenza vaccine
Conceptual model suggests that protection (and morbidity) could be peaked.
Model-based analysis of Norovirus vaccination strategies
Motivation
- For a future norovirus vaccine, explore what vaccination strategies have the most impact.
Methods
- Compartmental, SIR type ODE model.
- Model fit to norovirus incidence data.
Results
Tools
R packages to make modeling easier I
For teaching (stable):
R packages to make modeling easier II
For research (WIP):
Comments