Global First! Molecular Heart Xu Jinbo's team has opened up new algorithms to overcome the challenges of protein side chain prediction and sequence design

Tech 2023-06-05 15:22:13 Source: Network

Recently, the team led by Molecular Heart Xu Jinbo has launched a new AI protein side chain prediction algorithm called "AttnPacker", which has achieved significant improvements in speed, memory efficiency, and overall accuracy. It is currently the best known side chain structure prediction algorithm and the world's first AI algorithm that can simultaneously predict side chains and design protein sequences

Recently, the team led by Molecular Heart Xu Jinbo has launched a new AI protein side chain prediction algorithm called "AttnPacker", which has achieved significant improvements in speed, memory efficiency, and overall accuracy. It is currently the best known side chain structure prediction algorithm and the world's first AI algorithm that can simultaneously predict side chains and design protein sequences. The relevant paper has been published in the internationally authoritative academic journal "Proceedings of the American Academy of Sciences".

A protein is formed by folding several amino acids, and its structure is divided into main and side chains. The formation of protein structure and function largely depends on the interaction between side chain atoms. Therefore, accurate protein side chain prediction (PSCP) is a key link to solve the problem of protein structure prediction and protein design. When applied to the field of drug design, scientists can quickly and accurately find suitable binding sites for drugs and receptors, and even optimize or design binding sites as needed; In the field of enzyme optimization, scientists can optimize and modify protein sequences to involve multiple side chains in catalytic reactions, achieving more efficient and specific catalytic effects.

Professor Xu Jinbo is known as the "founder of AI protein folding" in the industry. In 2016, he proposed the first AI protein folding algorithm, which inspired Deepmind to develop AlphaFold, solving the problem of predicting the three-dimensional structure of most protein main chains. However, the prediction of protein side chain structure has not yet been completely broken through, which has become one of the important factors restricting the wide application of protein design technology in innovative drug research and development, synthetic biology and other fields.

Professor Xu Jinbo has been studying the prediction of protein side chain structure since 2003 and is one of the longest studied scientists in this field. His early papers on side chain structure prediction algorithms have been published at academic conferences and journals such as RECOMB, the top conference of international computational molecular biology, and Journal of the ACM, the official journal of the Association for Computing Machinery (ACM).

This time, Xu Jinbo's team innovatively developed AttnPacker, a new end-to-end deep learning method for predicting protein side chain coordinates. It combines simulation of side chain interactions and directly predicts side chain structures that are more physically feasible, with fewer atomic collisions and more ideal bond lengths and angles. AttnPacker has good predictive performance, high efficiency, and is more user-friendly, which is conducive to its widespread use in scientific research and industrial fields, "said Professor Xu Jinbo.

In terms of prediction performance, AttnPacker shows the best prediction accuracy and efficiency compared to previous state-of-the-art methods. Simultaneously ensuring physical authenticity. Due to the adoption of innovative methods. Compared with other methods, AttnPacker significantly improves computational efficiency and reduces inference time by more than 100 times.

AttnPacker performs equally well in protein design. Molecular Heart trained an AttnPacker variant for collaborative design. The protein sequence designed by this variant achieves key indicators such as natural sequence recovery rate, accuracy, and energy, which are comparable to or surpass the state-of-the-art protein design algorithms.

In addition to its amazing effectiveness and efficiency, AttnParker also has a very practical value - it is very easy to use and only requires a protein structure file to run.

To promote the development and application of this AI algorithm, Molecular Heart has opened up AttnPacker's pre training model, source code, and inference scripts on Github.

We will continue to explore more accurate and efficient protein prediction and design algorithms based on AI, "Xu Jinbo said, hoping that algorithms such as AttnPacker can further promote the solution of protein design needs in industrial applications.

Disclaimer: The content of this article is sourced from the internet. The copyright of the text, images, and other materials belongs to the original author. The platform reprints the materials for the purpose of conveying more information. The content of the article is for reference and learning only, and should not be used for commercial purposes. If it infringes on your legitimate rights and interests, please contact us promptly and we will handle it as soon as possible! We respect copyright and are committed to protecting it. Thank you for sharing.(Email:[email protected])

Mobile advertising space rental

Tag: Global First Molecular Heart Xu Jinbo team has opened