Live COVID-19 Cases
  • World 230,898,440
    World
    Confirmed: 230,898,440
    Active: 18,565,917
    Recovered: 207,599,369
    Death: 4,733,154
  • USA 43,404,877
    USA
    Confirmed: 43,404,877
    Active: 9,757,275
    Recovered: 32,947,854
    Death: 699,748
  • India 33,563,421
    India
    Confirmed: 33,563,421
    Active: 301,610
    Recovered: 32,815,731
    Death: 446,080
  • Brazil 21,283,567
    Brazil
    Confirmed: 21,283,567
    Active: 395,672
    Recovered: 20,295,538
    Death: 592,357
  • UK 7,530,103
    UK
    Confirmed: 7,530,103
    Active: 1,318,749
    Recovered: 6,075,733
    Death: 135,621
  • Russia 7,333,557
    Russia
    Confirmed: 7,333,557
    Active: 590,719
    Recovered: 6,542,213
    Death: 200,625
  • France 6,971,493
    France
    Confirmed: 6,971,493
    Active: 172,253
    Recovered: 6,682,931
    Death: 116,309
  • Spain 4,940,824
    Spain
    Confirmed: 4,940,824
    Active: 162,303
    Recovered: 4,692,436
    Death: 86,085
  • Italy 4,645,853
    Italy
    Confirmed: 4,645,853
    Active: 106,559
    Recovered: 4,408,806
    Death: 130,488
  • Germany 4,175,557
    Germany
    Confirmed: 4,175,557
    Active: 149,187
    Recovered: 3,932,600
    Death: 93,770
  • China 95,894
    China
    Confirmed: 95,894
    Active: 990
    Recovered: 90,268
    Death: 4,636
Generic selectors
Exact matches only
Search in title
Search in content
Generic selectors
Exact matches only
Search in title
Search in content
vaccine fairness

BY Olivia McDonald

Technology

Developments In Healthcare AI: Protein Predictions

The structures of almost all proteins made within the human body have been predicted using AI.

JULY 29  2021

Follow

Using a program called AlphaFold made by Google, AI Firm DeepMind’s researchers have managed to predict the structures of 350,000 proteins that belong to humans and other organisms such as mice, yeast and fruit flies. This development is outstanding and could fuel the discovery of new drugs to treat diseases.

Before this discovery, only a few thousand proteins made by the human genome had confirmed structures. Proteins are essential building blocks of any living organism and understanding their unique shapes and patterns is crucial for medical advancements thus demonstrating the immensity of this recent research by the London-based AI company.

Now DeepMind has said that they will release the structures of these proteins to the world on a free database. In the upcoming months, another 100 million protein structures will be revealed which will essentially be all proteins known to science.

To run the database, DeepMind is partnering with the European Molecular Biology Laboratory (EMBL), which is an international research institution that already hosts a large database of protein information.

The scientific breakdown

The human body creates instructions for making human proteins in our genomes which is the DNA contained in the nuclei of human cells. In totality, there are around 20,000 of these proteins expressed in the human genome and biologists refer to this full composition as the ‘proteome’.
Proteins are long ribbons of amino acids which are twisted in long, complex knots. In order to know the function of a protein, a scientist must know its shape which is extremely important when finding drugs to treat illnesses. Usually, it can take months or weeks to find a protein’s shape but with this new AI technology from AlphaFold, this process can take just two days. AlphaFold can predict shapes to the nearest atom, therefore, is suitable for drug development and good enough to enable researchers to understand the protein’s function despite the fact that the predictions are not 100% accurate.

Speaking to the BBC, Chief Executive and Co-Founder of DeepMind, Dr Demis Hassabis said:

‘We believe it’s the most complete and accurate picture of the human proteome to date. We believe this work represents the most significant contribution AI has made to advancing the state of scientific knowledge to date. And I think it’s a great illustration and example of the kind of benefits AI can bring to society. We’re just so excited to see what the community is going to do with this.’

Prof Edith Heard, from EMBL, also said:

‘This will be transformative for our understanding of how life works. That’s because proteins represent the fundamental building blocks from which living organisms are made.’

Future uses

Traditional methods to work out a protein’s structure include X-Rays and Cryogenic Electron Microscopy (Cryo-EM). These techniques are not only extremely difficult to conduct because they depend on trial and error which can take painstaking years of work but also cost a huge amount of money and resources as they rely on multi-million-pound specialised equipment. Therefore, DeepMind’s research has been ground-breaking because it has provided a solution to a 50-year-old challenge to biology. As protein structure predictions have become more efficient they have also become more economical which has reduced the cost associated with experimentation. It is hoped that with the use of AI, the quality of life for millions of patients around the world can be improved. In addition, aside from the fact that the research can be used to help develop drugs and treatments for diseases, the technology can also be used to help design future crops that can resist climate change and enzymes that can break down plastics that are polluting the environment.

Researchers have already begun to engineer bacteria to secrete proteins that will make waste more biodegradable and easier to process. PET is a strong plastic we commonly see used in items like bottles, however, these bottles can take hundreds of years to decompose. Through AI technology with protein structures, the modified enzyme which is known as PETase starts to break down the plastic in just a few days which could revolutionise recycling and allow for the more eco-friendly use of plastic. Thus, DeepMind’s Protein Structure breakthrough has positive implications for more than just the health sector, but also the environment too.

What else has AI technology achieved?

AI is a multifarious form of technology with many different purposes and a wide range of applications across industries. Some of its applications include the robotics industry, financial services and machine learning for wildlife conservation. But the most prominent advancements of AI technology have been in the health sector. AI has helped provide solutions to brain disorders like Parkinson’s disease with neuroelectronic technologies as well as facilitating COVID-19 vaccine production.

The Brookings Institution, a non-profit American public policy organisation, said that the aim of this vaccine production is to add strongly immunogenic viral components that caused the immune system to respond and machine learning has helped with this process. Experts have said that AI has the ability to identify viral components that have the necessary properties to stimulate the immune system. Thus, the technology has been important for researchers throughout the pandemic.

It is clear to see that AI has had an unbelievable impact on the world and has changed the health sector for the better. With these new technological advancements, the future looks very promising in fighting against diseases and discovering more about the human body with more efficiency.

 

About the Author: Olivia McDonald

Olivia McDonald is a contributing features writer, with a special interest in history and politics of the Americas. Her expertise lies in social and political issues that affect people from marginalised communities.

Recommended for you

Creativity Quantified: The Secret Behind Creative Hot Streaks

A study conducted with the help of deep-learning algorithms discovered that revered creatives have two things to thank for their hot streaks: exploration and exploitation.

Trending