The AI program AlphaFold2 can predict the key characteristics of uromodulin protein filaments. Its cryo-EM structure reveals a complex structure in which one subunit (blue) encloses the other two subunits (teal and magenta). The overall shape of the AI prediction is not the same as the cryo-EM model due to the lack of contact in the area indicated by the red arrow, but two halves accurately reproduce the experimental data (below). Panel). Credit: Luca Jovine
Scientists at the Karolinska Institutet report that machine learning can be used to gain insight into molecular events that change the shape of a protein after it is made and regulate its ability to interact. This suggests that artificial intelligence (AI) may allow us to accurately simulate highly complex biological scenarios in silico and use this information for therapeutic interventions in the future. I am.
Detailed knowledge of three dimensions (3D) Protein structure It is essential to understand their function, assess the effects of pathogenic mutations in humans, and support rational design of new drugs. recently, Machine learning The program AlphaFold2 has been shown to be able to predict this 3D information from. Protein sequence Alone with accuracy close to the experiment. Currently, Professor Luca Jovine of the Faculty of Bioscience and Nutrition at the Karolinska Institute says AlphaFold2 can also be applied to study molecular events that reshape proteins after they are synthesized and regulate their interaction with other molecules. I am reporting.
“This work, Computational approach The underlying related programs, such as AlphaFold 2 and RoseTTA Fold, can go far beyond accurate predictions of individual protein structures, “jovine said.
AI as an approach in structural biology
Determining the 3D structure of a protein is traditionally a long, difficult and costly process that relies on laboratory techniques such as X-ray crystallography, nuclear magnetic resonance (NMR), and more recently cryo-electron microscopy (EM). Was there. Against this background, the development of machine learning tools such as AlphaFold2 was quickly recognized as an important breakthrough in structural biology. However, the application of AI to more complex structural problems, such as understanding how proteins interact to form complexes, remains largely unexplored.
In this type of study, first published in Sweden, Jovine shows that AlphaFold2 can reproduce important parts of the complex molecular rearrangements that occur with uromodulin, a protein that protects our body. .. Urinary-tract infectionAfter being cleaved by another molecule, it is assembled into a filament. Given that AlphaFold2 had no knowledge of uromodulin filament structure, this surprising result is that the same basic approach is generally applicable to many other biomedically important molecular systems. Suggests. protein Synthetic.
The study was published in Molecular replication and development..
Luca Jovine uses machine learning to study protein-protein interactions: from uromodulin polymers to egg zona pellucida filaments Molecular replication and development (2021). DOI: 10.1002 / mrd.23538
Provided by
Karolinska Institute
Quote: Artificial intelligence takes structural biology to the next level (October 7, 2021). Obtained from https: //phys.org/news/2021-10-artificial-intelligence-biology.html on October 7, 2021.
This document is subject to copyright. No part may be reproduced without written permission, except for fair transactions for personal investigation or research purposes. The content is provided for informational purposes only.
Read more here: Source link