antigen-prime: Simulating coupled genetic and antigenic evolution of influenza virus
The paper introduces antigen-prime, a forward-time epidemic simulator that links genetic sequences to antigenic phenotypes under host selection to generate ground-truth data for benchmarking influenza variant assignment and growth rate estimation methods, revealing both their accuracy and specific failure modes.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of a preprint that has not been peer-reviewed. It is not medical advice. Do not make health decisions based on this content. Read full disclaimer
Influenza viruses are constantly changing their appearance to the human immune system. This process, known as antigenic drift, allows the virus to slip past the defenses built up by previous infections or vaccinations, leading to the seasonal outbreaks that affect millions each year. To predict which strains will dominate and to design effective vaccines, scientists try to group these viral variants based on how similar they look to our immune systems and how quickly they are spreading through a population. However, checking whether these grouping methods work correctly is notoriously difficult. In the real world, it is hard to measure exactly how well a virus can escape immunity or how fast it is growing at any given moment, leaving researchers without a reliable way to test their tools.
To solve this problem, a team of researchers developed a new computer program called antigen-prime. Instead of trying to measure the unpredictable real world, they built a virtual world where they could control every variable. This simulator acts like a time machine for the virus, running a forward-looking model of how influenza evolves over a period of thirty years. The key innovation is that this program links the virus's genetic code directly to its ability to evade the immune system, all while simulating the pressure of a host population trying to fight back. By creating this environment, the researchers generated a set of data where the true answers were known from the start, providing a perfect ground for testing other scientific methods.
Using this thirty-year simulation, the team first checked if their virtual virus behaved like the real thing. They found that the patterns of genetic change and immune escape in the simulation matched the patterns seen in nature, giving them confidence that the model was realistic. They then used this simulated data to test two different ways scientists currently assign viral variants to groups. One method relied solely on the genetic sequence, while the other looked at the family tree of the virus. The results showed that the method based on genetic sequences was slightly better at sorting the viruses into groups that were truly distinct from one another in terms of how the immune system sees them.
The researchers also tested how well current tools could estimate the growth rates of these viral groups over one-year periods. In most of the simulated years, the estimates were accurate, but in several specific windows, the tools failed dramatically. By looking closely at these moments of high error, the team uncovered a previously unreported way these methods can break down. This discovery highlights that while current models work well in many situations, they can miss critical shifts in viral behavior under certain conditions. The antigen-prime simulator itself is now available as an open tool for other scientists to use, offering a way to test and improve future models of how influenza evolves without waiting for the next natural outbreak to provide the answers.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.