AI Used to Predict 3D Structures of Proteins Made by Human Genome; Critical for Advancing Medicine

Scientists have used artificial intelligence (AI) to predict the structures of almost every protein in the human genome. This breakthrough technology has the potential to supercharge the discovery of new drugs to treat various diseases and shows promise in advancing medicine.

DeepMind and the European Molecular Biology Laboratory (EMBL), Europe's flagship laboratory for life science, collaborated to make a complete database of predicted protein structures for human proteome that covers approximately 20,000 proteins expressed by the human genome.
The database and AI system provide biologists with new tools to examine the 3D structure of proteins and offer a treasure trove of data that unlocks future advances in medicine.
They published the full findings of their study, titled "Highly accurate protein structure prediction for the human proteome," in the journal Nature.

(Photo: Pixabay)
AI Used to Predict 3D Structures of Proteins Made by Human Genome; Critical for Advancing Medicine.

AI Helps Scientists Accelerate Discovery

Scientists have used AlphaFold to predict 350,000 protein structures in the human genome and other organisms. These protein structures are found in the DNA of cells, wherein about 20,000 proteins are expressed by the human genome. This is collectively referred to as the human proteome, BBC News reported

However, predicting protein structures takes a lot of time, effort and uses costly techniques. But with the AI system, scientists found that predicting protein structures could be a lot easier now.

"It takes a huge amount of money and resources to do structures," structural biologist Professor John McGeehan of the University of Portsmouth told the news outlet.

DeepMind CEO and co-founder Dr. Demis Hassabis said that AlphaFold had created the most complete and accurate picture of human proteome so far, which proves that AI benefits society.

AlphaFold made complete and accurate 3D protein structures prediction for 58% of the amino acids in the human proteome. According to the news outlet, 35.7% were predicted with a very high degree of confidence.

In the past, each prediction of protein structure could take about six months, but AlphaFold can only do it for a couple of minutes. Professor McGeehan said that they could not have predicted this breakthrough and that it would happen so fast.

Applications of the new technology could accelerate the discovery of new drugs and treatments for disease. It could also be used to design future crops that can survive climate change or develop enzymes that break down plastics, which Professor McGeehan and his team are currently working on. They said that this accelerated their project by multiple years.

ALSO READ: How The Transcription Factor Guides the DNA in Specialized Cell Development

AlphaFold's Sophisticated AI System

According to EurekAlert!, AlphaFold's AI system, also known as AlphaFold Protein Structure Database, is built on contributions from the international scientific community, AlphaFold's sophisticated algorithmic innovations, and the decades of sharing the world's biological data of EMBL-EBI.

EMBL Deputy Director-General ad EMBL-EBI Director Ewan Birney said that AlphaFold would become the most important dataset since mapping the human genome, which makes predicting protein structures more accessible to the international scientific community to open doors to new scientific discoveries. It will be a great new scientific tool that complements technologies and shall pave the way to break boundaries of understanding the world.

The database also launches protein structures of other organisms, aside from human proteomes, such as E.Coli, fruit fly, mice, zebrafish, tuberculosis bacteria, and the parasite that causes malaria. These structures will hopefully allow researchers to cross a huge variety of fields from neuroscience to medicine to accelerate their projects and provide necessary technologies to society.