Warning: "continue" targeting switch is equivalent to "break". Did you mean to use "continue 2"? in /home/staffituk/public_html/wp-content/themes/Divi/includes/builder/functions.php on line 5941
Google’s DeepMind AI Makes Dramatic Breakthrough | Staff IT UK

Google’s DeepMind AI machine learning program has solved a genetic protein challenge, the result of will dramatically change biology and could bring many medical benefits.

The Challenge

Proteins are the building blocks of cells and, for more than half a century, scientists have been trying to find a way to accurately predict the 3D shape and the folds of protein structures from their amino-acid sequence.  This is because a protein’s function is intricately linked with its shape and, therefore, the ability to predict a protein’s structure is the key to the understanding of what it does and how it works. This area of study is called Critical Assessment of Structure Prediction (CASP).

Why?

Being able to do so will mean that the function of the thousands of ‘unsolved’ proteins in the human genome can be identified and disease-causing gene variations that differ between people can be better understood.  This could make it much faster for humans to develop medicines and treatments for more conditions, could have a huge impact on life sciences, and make it possible to study living things in new ways.

DeepMind AI

Google’s machine learning, AI network offshoot ‘DeepMind’ used a program called AlphaFold to beat 100 other teams trying to solve the same problem. The AlphaFold system achieved a median score of 92.4 GDT overall across all targets meaning that its predictions have an average error of only 1.6 Angstroms (the width of an atom).  The system was ‘trained’ using publicly available data consisting of around 170,000 protein structures from the protein data bank as well as large databases containing protein sequences of unknown structure.

The methods that have been closest to solving this long-running scientific challenge in recent years have been advanced X-ray crystallography and cryo-electron microscopy (cryo-EM) but DeepMind’s AlphaFold program produced results that took leading scientists in the field by surprise and were far ahead of those results by the nearest competing teams.

GameChanger

Andrei Lupas of the Max Planck Institute for Developmental Biology in Tübingen, Germany, who has been working on this same challenge for a decade, said: “This will change medicine. It will change research. It will change bioengineering. It will change everything”.

What Does This Mean For Your Business?

This is a huge success for Google and its DeepMind project and shows how concentrated power of AI can dramatically speed up the solving of problems that have eluded scientists for decades.  Not only is this a validation of Google’s AI efforts but also opens up opportunities for medical research and pharmaceutical companies, and could bring benefits to humanity in terms of being able to develop treatments for conditions that have offered little hope up until now. The pandemic has taught the world a powerful lesson in the importance of being able to develop fast ways of understanding new diseases and developing effective treatments and this is one area that DeepMind’s discovery could make a very positive contribution towards.