PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces
Abstract Proteins are essential molecular building blocks of life, responsible for most biological functions as a result of their specific molecular interactions. However, predicting their binding interfaces remains a challenge. In this study, we present a geometric transformer that acts directly...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2023-04-01
|
| Series: | Nature Communications |
| Online Access: | https://doi.org/10.1038/s41467-023-37701-8 |
| _version_ | 1797840899631218688 |
|---|---|
| author | Lucien F. Krapp Luciano A. Abriata Fabio Cortés Rodriguez Matteo Dal Peraro |
| author_facet | Lucien F. Krapp Luciano A. Abriata Fabio Cortés Rodriguez Matteo Dal Peraro |
| author_sort | Lucien F. Krapp |
| collection | DOAJ |
| description | Abstract Proteins are essential molecular building blocks of life, responsible for most biological functions as a result of their specific molecular interactions. However, predicting their binding interfaces remains a challenge. In this study, we present a geometric transformer that acts directly on atomic coordinates labeled only with element names. The resulting model—the Protein Structure Transformer, PeSTo—surpasses the current state of the art in predicting protein-protein interfaces and can also predict and differentiate between interfaces involving nucleic acids, lipids, ions, and small molecules with high confidence. Its low computational cost enables processing high volumes of structural data, such as molecular dynamics ensembles allowing for the discovery of interfaces that remain otherwise inconspicuous in static experimentally solved structures. Moreover, the growing foldome provided by de novo structural predictions can be easily analyzed, providing new opportunities to uncover unexplored biology. |
| first_indexed | 2024-04-09T16:22:22Z |
| format | Article |
| id | doaj.art-082cd95ef19544f996bec815b6afa158 |
| institution | Directory Open Access Journal |
| issn | 2041-1723 |
| language | English |
| last_indexed | 2024-04-09T16:22:22Z |
| publishDate | 2023-04-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Nature Communications |
| spelling | doaj.art-082cd95ef19544f996bec815b6afa1582023-04-23T11:22:07ZengNature PortfolioNature Communications2041-17232023-04-0114111110.1038/s41467-023-37701-8PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfacesLucien F. Krapp0Luciano A. Abriata1Fabio Cortés Rodriguez2Matteo Dal Peraro3Institute of Bioengineering, School of Life Sciences, Ecole Fédérale de Lausanne (EPFL) and Swiss Institute of Bioinformatics (SIB)Institute of Bioengineering, School of Life Sciences, Ecole Fédérale de Lausanne (EPFL) and Swiss Institute of Bioinformatics (SIB)Institute of Bioengineering, School of Life Sciences, Ecole Fédérale de Lausanne (EPFL) and Swiss Institute of Bioinformatics (SIB)Institute of Bioengineering, School of Life Sciences, Ecole Fédérale de Lausanne (EPFL) and Swiss Institute of Bioinformatics (SIB)Abstract Proteins are essential molecular building blocks of life, responsible for most biological functions as a result of their specific molecular interactions. However, predicting their binding interfaces remains a challenge. In this study, we present a geometric transformer that acts directly on atomic coordinates labeled only with element names. The resulting model—the Protein Structure Transformer, PeSTo—surpasses the current state of the art in predicting protein-protein interfaces and can also predict and differentiate between interfaces involving nucleic acids, lipids, ions, and small molecules with high confidence. Its low computational cost enables processing high volumes of structural data, such as molecular dynamics ensembles allowing for the discovery of interfaces that remain otherwise inconspicuous in static experimentally solved structures. Moreover, the growing foldome provided by de novo structural predictions can be easily analyzed, providing new opportunities to uncover unexplored biology.https://doi.org/10.1038/s41467-023-37701-8 |
| spellingShingle | Lucien F. Krapp Luciano A. Abriata Fabio Cortés Rodriguez Matteo Dal Peraro PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces Nature Communications |
| title | PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces |
| title_full | PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces |
| title_fullStr | PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces |
| title_full_unstemmed | PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces |
| title_short | PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces |
| title_sort | pesto parameter free geometric deep learning for accurate prediction of protein binding interfaces |
| url | https://doi.org/10.1038/s41467-023-37701-8 |
| work_keys_str_mv | AT lucienfkrapp pestoparameterfreegeometricdeeplearningforaccuratepredictionofproteinbindinginterfaces AT lucianoaabriata pestoparameterfreegeometricdeeplearningforaccuratepredictionofproteinbindinginterfaces AT fabiocortesrodriguez pestoparameterfreegeometricdeeplearningforaccuratepredictionofproteinbindinginterfaces AT matteodalperaro pestoparameterfreegeometricdeeplearningforaccuratepredictionofproteinbindinginterfaces |