New findings suggest combining experimental data about the appearance of the flu virus with its genetic code can improve the predictive power of supercomputer models to forecast future strains

Combining genetic and experimental data into models about the influenza virus can help predict more accurately which strains will be most common during the next winter, says a study published recently in eLife.

The models could make the design of flu vaccines more accurate, providing fuller protection against a virus that causes around half a million deaths each year globally.

Vaccines are the best protection we have against the flu. But the virus changes its appearance to our immune system every year, requiring researchers to update the vaccine to match. Since a new vaccine takes almost a year to make, flu researchers must predict which flu viruses look the most like the viruses of the future. {module INSIDE STORY}

The gold-standard ways of studying influenza involve laboratory experiments looking at a key molecule that coats the virus called haemagglutinin. But these methods are labor-intensive and take a long time. Researchers have focused instead on using computers to predict how the flu virus will evolve from the genetic sequence of haemagglutinin alone, but these data only give part of the picture.

"The influenza research community has long recognized the importance of taking into account physical characteristics of the flu virus, such as how haemagglutinin changes over time, as well as genetic information," explains lead author John Huddleston, a PhD student in the Bedford Lab at Fred Hutchinson Cancer Research Center and Molecular and Cell Biology Program at the University of Washington, Seattle, US. "We wanted to see whether combining genetic sequence-only models of influenza evolution with other high-quality experimental measurements could improve the forecasting of the new strains of flu that will emerge one year down the line."

Huddleston and the team looked at different components of virus 'fitness' - that is, how likely the virus is to thrive and continue to evolve. These included how similar the antigens of the virus are to previously circulating strains (antigens being the components of the virus that trigger an immune response). They also measured how many mutations the virus has accumulated and whether they are beneficial or harmful.

Using 25 years of historical flu data, the team made forecasts one year into the future from all available flu seasons. Each forecast predicted what the future virus population would look like using the virus' genetic code, the experimental data, or both. They compared the predicted and real future populations of flu to find out which data types were more helpful for predicting the virus' evolution.

They found that the forecasts that combined experimental measures of the virus' appearance with changes in its genetic code were more accurate than forecasts that used the genetic code alone. Models were most informative if they included experimental data on how flu antigens changed over time, the presence of likely harmful mutations, and how rapidly the flu population had grown in the past six months. "Genetic sequence alone could not accurately predict future flu strains - and therefore should not take the place of traditional experiments that measure the virus' appearance," Huddleston says.

"Our results highlight the importance of experimental measurements to quantify the effects of changes to virus' genetic code and provide a foundation for attempts to forecast evolutionary systems," concludes senior author Trevor Bedford, Principal Investigator at the Vaccine and Infectious Disease Division, Fred Hutchinson Cancer Research Center, Seattle, Washington. "We hope the open-source forecasting tools we have developed can immediately provide better forecasts of flu populations, leading to improved vaccines and ultimately fewer illnesses and deaths from flu."

Researchers at the Institute of Industrial Science at The University of Tokyo sifted through experimental data to probe the possibility that supercooled water has a liquid-to-liquid phase transition between disordered and tetrahedrally structured forms. They found evidence of a critical point based on the cooperative formation of tetrahedra, and show its minor role in water's anomalies. This work shows that water's special qualities--which are essential for life--originate predominantly from the two-state feature.

Liquid water is indispensable for life as we know it, yet many of its properties do not conform with the way other fluids behave. Some of these anomalies, such as water's maximum density at 4°C and its large heat capacity, have important implications for living organisms. The origin of these features has sparked fierce debates in the scientific community since the time of Röntgen.

Now, researchers at The University of Tokyo have utilized a two-state model that posits the dynamical coexistence of two types of molecular structures in liquid water. These are the familiar disordered normal-liquid structure and a locally favored tetrahedral structure. As with many other phase transitions, there may be a "critical point" at which the correlation between tetrahedra takes on a power-law form, which means there will no longer be any "typical" length scale.  {module INSIDE STORY}

Using supercomputer simulations of water molecules, along with a comprehensive analysis of experimental structural, thermodynamic, and dynamic data--including X-ray scattering, density, compressibility, and viscosity measurements--the researchers were able to narrow down where a critical point should be if it exists.

"If the formation of tetrahedral structures in liquid water is cooperative under these conditions, then a liquid-liquid phase transition with a critical point is possible," lead author Rui Shi says.

The team showed that this occurs around a temperature of -90°C and a pressure of about 1,700 atmospheres. Experiments in this range are exceedingly difficult: because the water is so far below its normal freezing, ice crystals can quickly form. However, samples can remain liquid in a metastable "supercooled" state at these very high pressures.

"We saw evidence that the critical point is real, but its effect is almost negligible in the experimentally accessible region of liquid water because it is too far from the critical point. This means that water's anomalies come from the two-state feature and not from criticality," senior author Hajime Tanaka says. The scientists anticipate that this project will lead to the convergence of the long debate on the origin of water's anomalies and more experimental research to access the second critical point of water.