New AI That Predicts Shape of Proteins Could Solve 50-Year Problem
Advances in Artificial Intelligence (AI) have been nothing short of revolutionary, allowing us to progress in complex problem-solving and making huge strides in efficiency. Now, AI is being used to solve a puzzle that has perplexed scientists and biologists for almost 50 years – predicting the shape of proteins.
Proteins are one of the crucial building blocks of life, and understanding their structure is a key part of understanding how they work and interact with other molecules. Knowing the structure of proteins can potentially revolutionize medical treatments, drug-discovery and even improve our understanding of Covid-19 and how to treat it.
Researchers at the University of California, Berkeley, have combined two artificial intelligence algorithms to predict the 3D shape of proteins. Published in Science Magazine, their paper explains how this AI algorithm accurately identifies and visualizes different shapes of proteins, allowing us to predict the structure of proteins at a much faster rate than ever before.
What Is Protein Structure?
Before we explain how this new AI works and why it is so revolutionary, it is important to have a basic understanding of protein structure. Proteins are essential to the activities that sustain life, such as the function of cells, the structure of organs, and the interactions between neurons.
Each cell in the body contains proteins, and each protein is made up of long chains of proteins, which can be unfolded and reshaped in many different ways, forming different structures. Eventually these proteins fold into their native ‘active structures’ – structures that are necessary for their correct functioning – and these active structures are then used in different ways inside the body.
The Reason We Need to Predict Protein Shape
Understanding protein shape is absolutely essential to understanding how they work, and how they interact with other molecules. For example, if a new drug is developed, scientists need to know the structure and shape of the proteins it will interact with, in order to understand how effective it will be and how to optimize it for further research.
In addition, understanding protein shape will give scientists greater insights into a variety of human diseases. By knowing the exact structure of proteins, scientists can pinpoint any variations from the normal version and better understand and potentially cure diseases.
How Does This AI Predict Protein Shape?
Essentially, the AI developed by UC Berkeley uses two algorithms – one that continuously generates virtual 3D protein structures, and another that evaluates these structures and chooses the most likely models of protein structures.
This process is repeated, until the AI finds the most viable structure and selects it as the best one. This is a revolutionary process, as it is much more efficient than the current protein structure algorithms, which rely on trial and error, generating hundreds or thousands of potential structures and then slowly narrowing them down until the most likely structure is identified.
How Accurate Is the AI?
In the paper published by Science Magazine, the accuracy of the AI was tested on 732 different proteins. The results show that the AI is far more successful at accurately predicting the structure of proteins than current existing protein structure prediction tools.
Benefits of This AI
There are a number of potential benefits associated with this AI-based approach to protein structure prediction. Firstly, it is much faster than traditional methods, meaning that it can quickly assess hundreds of different protein structures and identify the most viable structure.
This speed allows the AI to quickly identify the structure of many proteins that were beyond the reach of scientists and could take years to research using traditional methods.
In addition, the AI is incredibly versatile, meaning it can assess a wide variety of protein structures and accurately predict the structure of proteins that have never been studied before. This could open up new possibilities in medical treatments and drug discovery, as scientists can now quickly identify structures of proteins associated with diseases and develop treatments.
This new AI has the potential to revolutionize the way we study proteins, allowing us to identify and predict protein structures much quicker than ever before. It could open up a new world of possibilities in medical treatment and drug-discovery, as well as improve our understanding of viruses like Covid-19.
Overall, this AI represents a huge breakthrough in the field of protein structure prediction, solving a problem that has perplexed scientists for half a century.
