Last Updated on 22/11/2021 by Sunaina
Google’s parent company, Alphabet, said Thursday that a new Alphabet firm will apply artificial intelligence approaches for medication research. It will draw on the work of DeepMind, another Alphabet unit that has done significant work using AI to predict protein structure.
The new business, Isomorphic Laboratories, will capitalise on that achievement by developing tools to aid in the discovery of novel medications. DeepMind CEO Demis Hassabis will also be the CEO of Isomorphic, although the two businesses will remain independent and work on occasion, according to a spokeswoman.
For years, experts have pointed to AI as a means to accelerate and reduce the cost of developing new pharmaceuticals to treat a variety of ailments. AI might assist in searching through databases of candidate molecules to locate those that best match a certain biological target, or in fine-tuning proposed compounds. Over the last two years, hundreds of millions of dollars have been spent in firms developing AI technologies.
According to Hassabis, Isomorphic will attempt to develop models that can anticipate how medications will interact with the body. It may use DeepMind’s protein structure research to find out how numerous proteins interact with one another. Instead of developing its own medications, the corporation may market its models. According to a spokesman for The Verge, it will focus on creating collaborations with pharmaceutical businesses.
Developing and testing medications, on the other hand, may be a more difficult task than determining protein structure. For example, even if two proteins have physical structures that match together, it’s difficult to predict how well they’ll really cling together. A medication candidate that appears promising based on how it functions chemically may not always work when administered to an animal or a person. As scientist and journalist Derek Lowe pointed out in Science last summer, more than 90% of medications that make it to a clinical trial fail. The majority of the issues aren’t caused by a flaw at the molecular level.
DeepMind’s work and Isomorphic’s projected work might help break down certain research bottlenecks, but they aren’t a simple answer for the myriad obstacles of drug development. “The difficult, resource-draining process of completing the biochemistry and biological evaluation of, say, pharmacological activities” will continue, according to Helen Walden, a professor of structural biology at the University of Glasgow.
