DeepMind Predicts Structure Of Almost Every Known Protein

An artificial intelligence company claims that its programme has successfully predicted the structure of almost all proteins known to science, effectively resolving one of biology’s “grand challenges” and opening the door to novel insights and advancements in a range of fields, including medicine, food security, and climate science. Alphabet, the parent company of Google, revealed on Thursday that DeepMind, an AI company, has enlarged the AlphaFold program’s free public database to include more than 200 million protein structures.

According to DeepMind CEO Demis Hassabis in a news briefing, the enormous library now includes the “entire protein universe,” drawn from the sequenced genomes of practically every life on the world. The basic blocks of life are lengthy, intricate chains of amino acids called proteins. In order to get vital insights into how these chains function, scientists have long worked to unravel how these chains are gracefully twisted and folded into 3D structures. For example, knowing a protein’s precise form and the interactions between its different molecules can aid researchers in identifying prospective targets for medical therapies.

Read more:

According to DeepMind, AlphaFold’s updated database now contains protein structures for many plant, bacterial, animal, and other creatures. Hassabis stated in a blog post about the milestone that these upgrades provide “new options for academics to use AlphaFold to progress their work on vital challenges, like sustainability, food instability, and neglected illnesses.” “AlphaFold not only solved a 50-year grand challenge, but it also became the first major proof point of our founding thesis: that artificial intelligence can dramatically accelerate scientific discovery, and in turn advance humanity,” the author wrote. “AI could accurately predict the shape of a protein down to atomic accuracy, at scale and in minutes.”

2020 saw the release of AlphaFold, and DeepMind dazzled the scientific community with the publication of a database of structures that essentially covered every protein found in the human body. Thousands of newly predicted protein structures were included in the so-called AlphaFold Protein Structure Database, which was created in association with the European Molecular Biology Laboratory. According to Hassabis, academics from all across the world are already using the extensive data to analyse issues like antibiotic resistance and plastic waste.

For instance, scientists at the University of Portsmouth in the UK revealed in July 2021 that they are utilising the database to assist design enzymes for recycling specific kinds of single-use plastics. John McGeehan, head of the University of Portsmouth’s Centre for Enzyme Innovation, stated in a statement at the time, “AlphaFold provides us with an intriguing new library of templates to construct faster, more stable, and less expensive enzymes for recycling plastic.”

Hassabis stated that DeepMind is striving to increase the size of its database, with a focus on applications in the fields of drug development, basic biological research, climate science, quantum chemistry, and fusion .In his essay, he claimed that “AlphaFold represents a peek of the future and what can be achieved with computational and AI technologies applied to biology.”

Related Articles

Back to top button