MS02 - MEPI-01 Part 2 of 2

Scenario Modeling to Inform Public Policymaking (Part 2)

Monday, July 14 at 4:00pm in Salon 12

SMB2025 Follow

Note: this minisymposia has multiple sessions. The other session is: and Part-1.

Zhilan Feng (National Science Foundation), John W Glasser, The US Centers for Disease Control and Prevention (CDC)

Description:

Models describe observations, underlying processes, or both. Their utility depends on how well they reproduce observations, those to which they have been fit or others. In descriptive modeling, the parameters of functions with desirable properties are adjusted to minimize discrepancies between predictions and observations. The parameters of mechanistic models cannot be estimated from the observations to which their predictions will be compared. Why? Mechanistic models are hypotheses about the processes giving rise to observations. Fitting their parameters is tantamount to assuming that the underlying processes have been modeled correctly. Hypotheses are tested by comparing their predictions to independent observations. In public health, dynamic models purporting to describe the processes by which pathogens are transmitted among human hosts are simulated under counter-factual conditions to inform policy. Unless the underlying processes have been modeled correctly, their predictions are unreliable. The only way to know if the predictions of mechanistic models are reliable is to compare them to accurate independent observations. The operating characteristics of surveillance systems are rarely considered, if known. Different periods in time-series from such systems are not independent. Averaging the predictions of ad hoc ensembles of such models does not solve the problem. Their undeniable merits notwithstanding, neither do Bayesian methods.

Room assignment: Salon 12

Junling Ma

University of Victoria

"Assess the effectiveness of Contact Tracing during the early stage of a pandemic"

Contact tracing is an important intervention measure to control infectious diseases. We present a new approach that tracks contacts in a randomly mixed population, which allow us to precisely model the contact tracing process. The model resulting from this new approach allows us to study the effect of contact tracing and isolation of diagnosed patients on the control reproduction number and number of infected individuals. However, we found that case counts alone during an early stage of an outbreak before susceptible population have been depleted is not sufficient to identify key contact tracing parameters such as coverage probability (the fraction of contacts successfully tracked) and testing rate. We need the reason that a patient is tested for diagnosis, i.e., whether they are quarantined and showing symptom, or voluntarily tested due to symptom, or contact tracing while showing symptom. We then apply our model to estimate the effect of contact tracing on the basic reproduction number and epidemic size in Ontario, Canada.

Sen Pei

Columbia University

"Addressing the challenge of imperfect observation processes in epidemic modeling"

Mathematical models calibrated to infectious disease data are widely used to understand epidemic dynamics and inform public health policy. However, real-world surveillance data often suffer from limitations due to imperfect observation processes, posing significant challenges for accurate modeling and inference. In this talk, I will highlight key challenges in epidemic modeling arising from imperfect data, present several studies that address these issues, and discuss promising directions for future research.

Troy Day

Queens University

"Modelling the distribution of fitness effects of new mutations"

The distribution of fitness effects of new mutations is key to our understanding of many evolutionary processes. Theoreticians have developed several models to help understand the patterns seen in data and many of these models match the data extremely well. Here we argue that this excellent match between models and data provides very weak evidence for their explanatory power. To do so we prove that even randomly chosen models make predictions that are indistinguishable from empirical data. This suggests that other analyses are required to properly assess models for the distribution of fitness effects of mutations and we provide some suggestions as to how this might be done. This is joint work with Olivier Cotto.

Zhilan Feng

National Science Foundation

"Mechanistic models are hypotheses"

Science involves perceiving patterns (events that are repeated) in observations, hypothesizing causal explanations (underlying processes), and testing them. Mathematical models either describe or provide explanations for patterns. The equations of descriptive models have convenient mathematical properties while those of mechanistic ones correspond to processes. The parameters of descriptive models are fit to observations by choosing values that minimize discrepant predictions. Because mechanistic models are hypotheses about the processes underlying patterns, their parameters should not be fit, but rather, based insofar as possible on first principles or estimated independently. The precision of mathematics facilitates comparing the predictions of mechanistic models to the patterns that they purport to explain and, until concordant, identifying and remedying the cause(s) of disparities.

#SMB2025 Follow

Annual Meeting for the Society for Mathematical Biology, 2025.