AlphaFold database adds AI-predicted protein structures for more than 2,800 viruses.
The AlphaFold Database now provides AI-Predicted Protein Structures for more than 2,800 viruses. Openly available to researchers for better understanding and protection against emerging pathogens. This dataset primarily focuses on proteomes, the family of viruses that are known to affect humans. These structural predictions, which are openly accessible, could help researchers investigate and identify reasons for viruses that surface and interact with human cells. The visualization of these 3D protein structures can help us with better diagnosis of therapeutics and vaccines. This information can also support researchers in being prepared for future outbreaks or pandemics like the COVID-19 outbreak.
An international collaboration among EMBL’s European Bioinformatics Institute (EMBL-EBI), Google DeepMind, NVIDIA, Seoul National University, the University of Glasgow, the Swiss Institute of Bioinformatics (SIB), the Coalition for Epidemic Preparedness Innovations (CEPI), and Sungkyungwan University in South Korea & Lund University in Sweden made this work possible.
The AI-predicted AlphaFold database also covers viruses that are less studied and is thus helping scientists with limited resources to confront future outbreaks first.
AI-Predicted Protein Structures: Why They Matter
Viruses enter the host cell using specific proteins. Understanding the 3D structures of these proteins will help us recognize how the virus will interact with the host cells. It can also help us understand and reveal the virus’s surface regions that can be key targets for diagnostics, therapeutics, and vaccines. Such information helps scientists to respond quickly if there is a possibility of a future pandemic.
“Making this data open is critical for understanding viral diagnostics and developing treatments and vaccines,” said JC McEntyre, interim director of EMBL-EBI.
These databases can support more effective and faster responses to any possible future pandemics, and they also cover the lesser-known or less-studied viruses to benefit a low-resource scientist.
How can this help us to prepare for a future pandemic?
The COVID-19 pandemic showed us the real awareness of how unprepared we are to prevent any global health emergencies. After an analysis by the Center for Global Development, it is estimated that there is a 50% chance of the world facing a pandemic that would be as severe as COVID-19 by the year 2050.
Preparation for such an outbreak in the future will be easier if we have knowledge about the viral family.
International events are called the 100 Days Mission, Aims to make the diagnostics, therapeutics, and vaccines available easier after the identification of another pandemic.
Having the structural knowledge of viruses, even the less-studied ones, will help researchers skip one of the first steps, and Scientists can make further predictions to decide which protein or molecular structure needs more research.
These AI-Predicted Protein Structures in the AlphaFold Database make it very useful to real-world science. Researchers everywhere, especially in regions where they face sudden disease outbreaks.
Drawbacks of these predictions
While these predicted protein structures show how the virus protein looks and how it might interact, they do not predict the genetic variations in a virus or a host-pathogen interaction. When a virus gets into a host cell, we cannot predict how structural changes make a virus more deadly or transmissible. For that, it requires an experimental investigation in the lab.
“A Protein complex structure alone does not tell us what happens when a virus mutates.”, explained Joe Grove, Professor of Molecular Virology at the MRC University of Glasgow Centre for Virus Research.
Open AlphaFold data for future pandemic preparation.
“Pandemic preparedness depends on having high-quality molecular information before an outbreak begins,” said Anthony Costa, director of digital biology at NVIDIA.
The open AlphaFold data coincides with the meeting that was held by the United Nations General Assembly on pandemic prevention, preparedness, and response, which had taken place on September 25 at the UN headquarters in New York.
AI-Predicted Protein Structures can be accessed worldwide by various researchers through the pandemic preparedness portal on the AlphaFold database. This demonstrates how powerful the combination of AI models and accurate data generated through human expertise can be in helping to address global health threats.
