A new AI system can warn us of next deadly virus in advance

The researchers tested EWAD on real data from the COVID-19 pandemic and found that it was able to accurately predict which variants of concern (VOCs) would arise as the virus mutated. The system could also estimate how public health measures such as vaccines and mask-wearing would affect the evolution of the virus.

The study, published in the journal Cell Patterns, shows that EWAD could help us prepare for and respond to future outbreaks by identifying potential threats before they are officially designated by the World Health Organization (WHO).

“We could see key gene variants appearing and becoming more prevalent, as the mortality rate also changed, and all this was happening weeks before the VOCs containing these variants were officially designated by the WHO,” says William Balch, a microbiologist at Scripps Research and one of the lead authors of the study.

The AI system uses a mathematical technique called Gaussian process-based spatial covariance, which can predict new data based on existing data and their relationships. The system can also detect patterns and rules of virus evolution that are otherwise hidden in the vast amount of data.

Lessons learned

“One of the big lessons of this work is that it is important to take into account not just a few prominent variants, but also the tens of thousands of other undesignated variants, which we call the ‘variant dark matter,’” says Balch.

The researchers say that their system could also help us understand more about the basic biology of viruses and how they adapt to different environments. This could lead to better treatments and prevention strategies for viral diseases.

“This system and its underlying technical methods have many possible future applications,” says Ben Calverley, a mathemologist at Scripps Research and another lead author of the study.

The study has been published in the journal Cell Patterns.

Study abstract:

We have developed a machine learning (ML) approach using Gaussian process (GP)-based spatial covariance (SCV) to track the impact of spatial-temporal mutational events driving host-pathogen balance in biology. We show how SCV can be applied to understanding the response of evolving covariant relationships linking the variant pattern of virus spread to pathology for the entire SARS-CoV-2 genome on a daily basis. We show that GP-based SCV relationships in conjunction with genome-wide co-occurrence analysis provides an early warning anomaly detection (EWAD) system for the emergence of variants of concern (VOCs). EWAD can anticipate changes in the pattern of performance of spread and pathology weeks in advance, identifying signatures destined to become VOCs. GP-based analyses of variation across entire viral genomes can be used to monitor micro and macro features responsible for host-pathogen balance. The versatility of GP-based SCV defines starting point for understanding nature’s evolutionary path to complexity through natural selection.