A new artificial intelligence tool called EVEscape has been developed by researchers at Harvard Medical School and the University of Oxford to predict the emergence of viral variants. By using evolutionary and biological information, EVEscape can forecast how a virus may change to evade the immune system. The tool has two key elements: a model of evolutionary sequences that predicts possible changes in a virus, and detailed biological and structural information about the virus. These combined elements enable EVEscape to make predictions about the most likely variants to occur as the virus evolves.
In a study published in the journal Nature, researchers demonstrated that if EVEscape had been deployed at the start of the COVID-19 pandemic, it would have accurately predicted the most frequent mutations and identified the most concerning variants of the SARS-CoV-2 virus. The tool also showed accurate predictions for other viruses, including HIV and influenza.
EVEscape is now being used to forecast future variants of SARS-CoV-2 every two weeks, with the researchers releasing a ranking of new variants. The information provided by EVEscape could eventually assist scientists in developing more effective vaccines and therapies for SARS-CoV-2 and other rapidly mutating viruses. The researchers are also expanding their work to include more viruses.
Senior author Debora Marks, associate professor of systems biology at Harvard Medical School, stated, “We want to know if we can anticipate the variation in viruses and forecast new variants because if we can, that’s going to be extremely important for designing vaccines and therapies.”
EVEscape is a modified version of the evolutionary model of variant effect (EVE) initially developed by the same researchers to predict gene mutations causing human diseases. The core of EVE is a generative model that can predict protein functionality based on evolutionary data across species. The researchers recognized the opportunity to adapt EVE to predicting viral variants amidst the COVID-19 pandemic. They combined the generative model from EVE with biological and structural details about the virus, including information about regions targeted by the immune system. This flexible framework allows EVEscape to be easily adapted for any virus.
The study involved predicting the mutations that would occur in the SARS-CoV-2 virus during the early stages of the pandemic, with EVEscape demonstrating similar accuracy to experimental approaches in predicting the virus’s ability to bind to antibodies. Not only did EVEscape outperform these approaches in predicting the most prevalent mutations, but it also produced quicker and more efficient predictions since it did not rely on waiting for relevant antibodies for testing.
EVEscape also identified which antibody-based therapies would lose their efficacy as the virus developed mutations to evade these treatments. Furthermore, the tool can analyze tens of thousands of new SARS-CoV-2 variants produced each week and pinpoint the ones most likely to become problematic.
The researchers believe that EVEscape can revolutionize vaccine development, aid in the surveillance of potential pandemic viruses, and facilitate quicker public health decision-making. They are currently using the tool in real-time to predict the future evolution of SARS-CoV-2 and publish a biweekly ranking of new variants. The team is also testing EVEscape on understudied viruses such as Lassa and Nipah.
EVEscape has the potential to design mutation-proof vaccines and therapies by evaluating their effectiveness against current and future viral variants. This proactive approach can help scientists develop treatments that can withstand the escape mechanisms acquired by the virus.
1. Source: Coherent Market Insights, Public sources, Desk research
2. We have leveraged AI tools to mine information and compile it