Google's DeepMind just shared AI-generated predictions about the coronavirus that could help researchers stem the global outbreak

Google's DeepMind just shared AI-generated predictions about the coronavirus that could help researchers stem the global outbreak

Demis Hassabis

Kim Hong-Ji/Reuters

DeepMind CEO Demis Hassabis.

  • DeepMind, Google's AI unit, just published predictions of the protein structures associated with the coronavirus that causes the disease known as COVID-19.
  • These predictions were drawn from DeepMind's new deep learning system but have yet to be experimentally verified, DeepMind noted in a blog post announcing their findings.
  • DeepMind said that it would ordinarily wait for findings to be reviewed by an academic journal, but it's skipping that process given the "potential seriousness and time sensitivity of the situation."
  • The predictions are open-sourced, allowing any researcher to build on, adapt or share DeepMind's findings.
  • Visit Business Insider's homepage for more stories.

DeepMind, Google's AI unit, is joining the global effort learn more about COVID-19, as the disease's toll spreads rapidly throughout much of the world.

The UK-based team just used DeepMind's new deep learning system to share its predictions about the protein structures of the coronavirus, which causes the disease. The system uses a machine-learning technique known as "free modelling" to help it predict protein structures when no similar structures of protein are available.
Advertisement
Scientists around the world are racing to learn more about the new coronavirus that has swept through China and spread across 86 other countries, at last count. At least 288 people have died outside of mainland China, as of Thursday.

DeepMind's findings hope to cut down on months of effort typically used to determine the protein structure of a virus.

"Knowing a protein's structure provides an important resource for understanding how it functions, but experiments to determine the structure can take months or longer, and some prove to be intractable," a blog post announcing the findings noted.
TOP VIDEOS FOR YOU
Advertisement

And given the "potential seriousness and time sensitivity of the situation," DeepMind said that it is skipping the process of experimentally verifying the findings or waiting for it to be peer-reviewed by an academic journal before publication. That's in line with other scientific research that has been appearing in both in peer-reviewed journals and on pre-print servers without peer review, since that process can take months.

"We emphasise that these structure predictions have not been experimentally verified, but hope they may contribute to the scientific community's interrogation of how the virus functions, and serve as a hypothesis generation platform for future experimental work in developing therapeutics," the blog post said. "Normally we'd wait to publish this work until it had been peer-reviewed for an academic journal. However, given the potential seriousness and time-sensitivity of the situation, we're releasing the predicted structures as we have them now, under an open license so that anyone can make use of them," it added.
Advertisement

The open license will allow any researcher to build on, adapt or share DeepMind's findings.

Google acquired the London-based research entity DeepMind for 400 million pounds back in 2014. The company has previously used AI to push advances in healthcare, developing models for identifying eye diseases and detecting neck cancer.

{{}}