The limitations of non-mechanistic methods for characterizing pathogen-pathogen interactions: A simulation study
Sarah C. Kramer, Sarah Pirikahu, Cana Kussmaul, Lulla Opatowski, Matthieu Domenech de Cellès
Abstract
Pathogen-pathogen interactions occur when infection with one pathogen influences one’s chance of infection or disease due to another. Increasingly, evidence suggests that interactions are a common feature of infectious disease epidemiology. However, due to both the nonlinearities and stochasticity inherent to infectious disease transmission, and the frequency of confounding (e.g., by shared seasonal forcing), simple, correlative methods for characterizing interactions may be prone to failure.
Introduction
Although infectious diseases are most commonly studied in isolation, pathogen-pathogen interactions, in which infection with one pathogen modifies the risk of infection or disease due to another pathogen, are rapidly gaining more attention. Throughout the literature, the term “interaction” is applied broadly and sometimes inconsistently: population-level phenomena (e.g., reductions in cases of subsequent pathogens due to people remaining at home during convalescence) and even purely correlative relationships are sometimes included.
Methods
Implementation
All analyses were conducted in R version 4.4.0 [41]. All code used for this project has been published on GitHub [42].
Results
Symmetric interactions between pathogens led to observable differences in epidemic dynamics
Changing the strength and duration of the interaction between our two modeled pathogens led to observable differences in the dynamics of both pathogens, particularly for long-lasting interactions (Figs 2A and S1). This suggests that outbreak data, which are collected at the population level, may nonetheless contain signal of individual-level interactions. When holding all other model parameters constant, negative interactions generally reduced both the average attack rate and the average peak week of pathogen B.
Discussion
Although pathogen-pathogen interactions are common, they are notoriously difficult to study in human populations. Due to the nonlinear dynamics underlying infectious disease transmission, the pervasiveness of confounding, and the inherent complexity of interactions themselves, many simple statistical methods fail when applied to questions about interactions [19, 20]. Here, we tested five non-mechanistic statistical methods and evaluated their ability to identify and characterize the interaction between two pathogens using synthetic data.
Acknowledgments
The authors gratefully acknowledge Anabelle Wong, Pietro Gemo, and Franziska Frederking for their helpful comments on the manuscript.
Citation: Kramer SC, Pirikahu S, Kussmaul C, Opatowski L, Domenech de Cellès M (2026) The limitations of non-mechanistic methods for characterizing pathogen-pathogen interactions: A simulation study. PLoS Comput Biol 22(8): e1013859. https://doi.org/10.1371/journal.pcbi.1013859
Editor: Burcu Tepekule, Princeton University Department of Ecology and Evolutionary Biology, UNITED STATES OF AMERICA
Received: December 21, 2025; Accepted: July 22, 2026; Published: August 17, 2026
Copyright: © 2026 Kramer et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: All code necessary to reproduce the findings reported here is available in our Git repository: https://github.com/sarahckramer/Pathogen_interaction_simulations (DOI: 10.5281/zenodo.17791160). The synthetic data used for the majority of the analyses can be reproduced using the code linked above. Data on influenza and RSV from Hong Kong and Canada used for the analysis of real-world data can be found at https://www.chp.gov.hk/en/statistics/data/10/641/642/2274.html and https://search.open.canada.ca/opendata/?sort=metadata_modified+desc&search_text=fluwatch&page=1, respectively. Climate data from the US National Centers for Environmental Information can be downloaded using the R package GSODR.
Funding: The author(s) received no specific funding for this work.
Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: MddC received consulting fees from MSD, GSK, Moderna, and Vaxcyte for work unrelated to this project. LO received a research grant from Sanofi for work on pathogen-pathogen interactions. All other authors declare no competing interests.
