Back

Drivers of COVID-19 variant wave dynamics: inferring oncoming wave size using global data with genomics

Molan, S.; Smith, N. K.; Gandhi, V.; Li, M.; Colijn, C.; Murall, C. L.; Stockdale, J. E.

2025-09-18 epidemiology
10.1101/2025.09.16.25335896 medRxiv
Show abstract

The continued evolution of the SARS-CoV-2 virus drove waves of infection worldwide throughout the pandemic. These evolutionary dynamics posed significant challenges for public health forecasting and, specifically, for predicting the size of COVID-19 waves. In this work we leverage a range of global public data, with a focus on features derived from pathogen genomic sequences, to model and predict the relative size of COVID-19 waves (as compared to the previous wave) across countries. Focusing on Omicron BA.1 and BA.2, we develop statistical models to assess the predictive power of these data in forecasting future variant-driven wave peaks. We find that, while forecasting wave size is a challenging task, variables such as genomic variant characteristics, prior wave dynamics, and demographic features e.g. life expectancy were informative, whereas seasonality was not. Our results show that the importance of features changed markedly between Omicron waves, reflecting the evolving epidemiological and genomic landscape. This work provides insights into improving predictive models for future outbreaks and pandemics, and prioritizing data collection efforts to enhance forecasting accuracy.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.