Viral Evolution Data Improves Predictions of Flu S

Viral Evolution Data Improves Predictions of Flu Season Severity

NEW YORK (GenomeWeb) – By folding in genetic data on how influenza viruses are evolving, researchers can better forecast how bad a coming flu season will be.

The seasonal flu accounts for some 1 billion illnesses a year and for 250,000 to 500,000 deaths a year worldwide. Four strains — H3N2, H1N1, and two influenza B viruses — circulate annually, with slight genetic variations from year to year. As they reported today in Science Translational Medicine, University of Chicago researchers have now taken this genetic evolution into account, alongside other data, in a new model to predict H3N2 incidence.


"Combining information about the evolution of the virus with epidemiological data will generate disease forecasts before the season begins, significantly earlier than what is currently possible," senior author Mercedes Pascual, professor of ecology and evolution at UChicago, said in a statement.

She and her colleagues first tweaked an epidemiological model to explain monthly flu incidence data collected in the US between October 2002 and June 2016. A basic model following the susceptible-infected-recovered-susceptible formulation that divvies the population into immune, infected, and non-immune individuals could reproduce the average seasonality of H3N2 flu incidence, but not variation from year to year, the researchers reported.

By adding H1N1 flu incidence as an influence on H3N2 flu, the researchers' model improved. And additional parameter fine-tuning such as adding in variability in reporting outside the prime flu season refined the epidemiological model even further.

The researchers also developed an index of viral evolutionary change based on alterations to hemagglutinin that occurred between 2002 and 2016. These protein sequences were obtained from the Global Initiative on Sharing Avian Influenza Data.

Changes to viral antigens like hemagglutinin could make people who were exposed to previous iterations of the flu susceptible to the next round, if the shift is large enough. They compared each year's sequences to gauge the extent of the change.

"Every two or three years, there is a big genetic change in the virus, which can make many more people sick," Xiangjun Du, a postdoc at UChicago, said in the statement. "Without factoring evolution into the model, you cannot capture these peaks in the number of cases."

Current forecasting models, the researchers noted, rely solely on mathematical models of how quickly the disease is spreading.

The UChicago researchers trained their model on that flu data collected between 2002 and 2011. They then used it to predict the severity of the 2011 to 2016 flu seasons. They reported that their approach accurately predicted whether those five seasons would have high or low severity. While their model did tend to underestimate the incidence peaks, they noted that it did capture the overall trend from year to year.

They also used the data available through the end of June 2016 to make predictions about the 2016 to 2017 flu season. They forecast a high-risk season for last fall before it began.

But what about this season? "That's the million-dollar question," Pascual said. "Our analysis for this year showed that the virus is already changing in a significant way. We predict an outbreak that is above average, but moderate, not severe, because last year was such a bad season."

Future versions of the model could take other factors into considerations, such as changes to other viral genes, age and social structure of the human population, and vaccination dynamics, the researchers noted.


https://www.genomeweb.com/infectious-disease/viral-evolution-data-improves-predictions-flu-season-severity

Research Article: http://stm.sciencemag.org/content/9/413/eaan5325