Article
Biochemistry & Molecular Biology
Dong Si, Jason Chen, Andrew Nakamura, Luca Chang, Haowen Guan
Summary: The study of macromolecular structures has revolutionized the pharmaceutical industry's development of new vaccines. While X-ray crystallography has been the traditional method, cryo-EM is gaining popularity due to recent advancements. Automated software is desired for labor-intensive de novo macromolecular complex modeling. This paper presents deep learning-based methods for macromolecule modeling applications.
JOURNAL OF MOLECULAR BIOLOGY
(2023)
Article
Biochemistry & Molecular Biology
Bahareh Behkamal, Mahmoud Naghibzadeh, Mohammad Reza Saberi, Zeinab Amiri Tehranizadeh, Andrea Pagnani, Kamal Al Nasr
Summary: Cryo-electron microscopy (cryo-EM) is a significant technique for protein structure determination, but limited by its resolution. This study proposes a novel automatic computational method to identify secondary structure elements (SSEs) in a protein's 3D structure, demonstrating efficiency and robustness in experimental evaluation.
Article
Computer Science, Artificial Intelligence
Shigeyuki Matsumoto, Shoichi Ishida, Mitsugu Araki, Takayuki Kato, Kei Terayama, Yasushi Okuno
Summary: Cryo-electron microscopy (cryo-EM) is a powerful tool for determining protein structures at atomic-level resolution. The DEFMap deep learning approach can extract dynamics hidden in cryo-EM density maps, providing valuable insights into protein function and structural changes. This multidisciplinary approach combines experimental data, molecular dynamics simulations, and deep learning techniques to study protein dynamics.
NATURE MACHINE INTELLIGENCE
(2021)
Article
Biology
Jonathan Bouvette, Qinwen Huang, Amanda A. Riccio, William C. Copeland, Alberto Bartesaghi, Mario J. Borgnia
Summary: This article introduces SmartScope, a framework for streamlining, standardizing, and automating the evaluation of cryo-EM specimens. Through deep-learning-based object detection, SmartScope can automatically identify and classify features suitable for imaging, enabling fully automated specimen screening. It provides a web interface for remote real-time control of microscope operation and access to images and annotation tools. Manual annotations can be used to re-train the feature recognition models for improved performance. This automated tool simplifies structure determination and lowers the barrier of adoption for cryo-EM.
Article
Engineering, Electrical & Electronic
Harshit Gupta, Michael T. McCann, Laurene Donati, Michael Unser
Summary: CryoGAN is a new paradigm for single-particle cryo-electron microscopy reconstruction based on unsupervised deep adversarial learning, achieving 3D structure reconstruction via distribution matching and skipping pose estimation or marginalization. It has shown success in recovering correct structures in an idealized mathematical model and achieving 10.8 angstrom resolution on a realistic synthetic dataset.
IEEE TRANSACTIONS ON COMPUTATIONAL IMAGING
(2021)
Article
Biochemistry & Molecular Biology
Wim J. H. Hagen
Summary: Researchers have presented an energy filter-based plasmon imaging method for rapid screening of gold foil grids, in order to save expensive imaging time.
FRONTIERS IN MOLECULAR BIOSCIENCES
(2022)
Article
Computer Science, Artificial Intelligence
Fatma Gouizi, Ahmed Chaouki Megherbi
Summary: This paper proposes a deep nested background subtraction algorithm based on residual micro-autoencoder blocks, which can extract significant multi-scale features of a complex scene. Experimental results demonstrate that the proposed method outperforms other state-of-the-art methods on benchmark datasets.
INTERNATIONAL JOURNAL OF MULTIMEDIA INFORMATION RETRIEVAL
(2023)
Article
Chemistry, Multidisciplinary
Carsten E. Seyfert, Christoph Porten, Biao Yuan, Selina Deckarm, Fabian Panter, Chantal D. Bader, Janetta Coetzee, Felix Deschner, Kamaleddin H. M. E. Tehrani, Paul G. Higgins, Harald Seifert, Thomas C. Marlovits, Jennifer Herrmann, Rolf Mueller
Summary: In recent decades, the pipeline for antibiotics targeting Gram-negative bacteria has become depleted due to the insufficiency in potency, pharmacokinetic properties, and toxicity of most discovered candidates. A new class of promising drug candidates called darobactins has been developed, which bind to a novel antibiotic target called BamA, an outer membrane protein. This study presents the cryo-EM structures of the Bam complex with darobactin 9 (D9) and the biotechnological generation of twenty new darobactins, including halogenated analogs, using an optimized purification method. The newly engineered darobactin 22 binds more effectively to BamA and exhibits 32-fold improved activity against clinically relevant pathogens such as carbapenem-resistant Acinetobacter baumannii, without any observed toxic effects.
ANGEWANDTE CHEMIE-INTERNATIONAL EDITION
(2023)
Article
Chemistry, Multidisciplinary
Erney Ramirez-Aportela, Jose M. Carazo, Carlos Oscar S. Sorzano
Summary: Single-particle cryo-electron microscopy is a powerful technique for determining the 3D structure of biological molecules. This article explores the use of deepEMhancer, a deep-learning approach, as a regularizer in the refinement process. The combination of deepEMhancer with RELION improves image quality and is particularly useful for studying membrane proteins.
Article
Chemistry, Multidisciplinary
Timothy Wiryaman, Navtej Toor
Summary: Protein nanocompartments are common structures in bacteria and archaea, with functions that are not yet fully understood. The high-resolution cryo-EM structure of a nanocompartment from Thermotoga maritima reveals that interactions in the E-loop domain may be crucial for its thermostability. Additionally, channels at various axes are analyzed for their iron transport capabilities, and an unexpected flavin ligand on the shell suggests a potential direct role in iron metabolism.
Article
Virology
Ivy Wang, Sandeep K. Gupta, Guillaume Ems, Nadishka Jayawardena, Mike Strauss, Mihnea Bostina
Summary: The cryo-EM structure of the non-human enterovirus EV-F4 provides insights into the structural diversity of these picornaviruses. The absence of charged residues on the outer surface suggests a different receptor type than for EV-F3. This study highlights the importance of understanding the structure of enteroviruses for global health.
Article
Biochemistry & Molecular Biology
Blake T. Riley, Stephanie A. Wankowicz, Saulo H. P. de Oliveira, Gydo C. P. van Zundert, Daniel W. Hogan, James S. Fraser, Daniel A. Keedy, Henry van den Bedem
Summary: The new qFit 3 software automates the analysis of dynamic protein structure data by generating a multi-conformer model, providing important insights for biological research. The software has been improved to run efficiently on small servers, desktops, or laptops. qFit 3 is open-source and available at the specified URL.
Article
Biochemistry & Molecular Biology
Maytha Alshammari, Jing He
Summary: The study aims to enhance the accuracy of ranking secondary structure topologies by combining three sources of information. Significant long-range contact pairs are found to be most effective in enriching the rank of the maximum-match topology for proteins with a large number of secondary structures. The quality of initial topology candidates is observed to have a significant impact on the enrichment of the approach.
Article
Multidisciplinary Sciences
Yanting Tang, An Mu, Yuying Zhang, Shan Zhou, Weiwei Wang, Yuezheng Lai, Xiaoting Zhou, Fengjiang Liu, Xiuna Yang, Hongri Gong, Quan Wang, Zihe Rao
Summary: Encapsulins containing dye-decolorizing peroxidase (DyP)-type peroxidases play a role in protecting prokaryotic cells against oxidative stress. Through isolating and studying a native cargo-packaging encapsulin from Mycobacterium smegmatis, researchers have determined the high-resolution structure and proposed a potential mechanism for hydrogen peroxide removal. The study suggests DyP as the primary cargo protein of mycobacterial encapsulins and a potential target for antituberculosis drug discovery.
PROCEEDINGS OF THE NATIONAL ACADEMY OF SCIENCES OF THE UNITED STATES OF AMERICA
(2021)
Article
Biochemical Research Methods
Afshin Khadangi, Thomas Boudier, Vijay Rajagopal
Summary: This research introduces a platform called EM-stellar for benchmarking the performance of various state-of-the-art DL methods on user-provided datasets. The study finds that the performance of any DL method depends on the properties of the images being segmented, and no single DL method consistently performs across all performance evaluation metrics.
Article
Biochemical Research Methods
Zeyu Wen, Jiahua He, Huanyu Tao, Sheng-You Huang
Article
Biochemical Research Methods
Jiahua He, Huanyu Tao, Sheng-You Huang
Article
Biochemistry & Molecular Biology
Jiahua He, Jun Wang, Huanyu Tao, Yi Xiao, Sheng-You Huang
NUCLEIC ACIDS RESEARCH
(2019)
Article
Biochemical Research Methods
Zeyu Wen, Jiahua He, Sheng-You Huang
Article
Biochemistry & Molecular Biology
Yumeng Yan, Jiahua He, Yuyu Feng, Peicong Lin, Huanyu Tao, Sheng-You Huang
PROTEINS-STRUCTURE FUNCTION AND BIOINFORMATICS
(2020)
Article
Biochemical Research Methods
Yumeng Yan, Huanyu Tao, Jiahua He, Sheng-You Huang
Article
Virology
Jiahua He, Huanyu Tao, Yumeng Yan, Sheng-You Huang, Yi Xiao
Article
Biochemical Research Methods
Jiahua He, Sheng-You Huang
Summary: The study presents a semi-automatic de novo structure determination method named DeepMM that can build atomic-accuracy all-atom models from cryo-EM maps at near-atomic resolution. It uses Densely Connected Convolutional Networks to predict main-chain, Ca positions, amino acid, and secondary structure types in the EM map. DeepMM showed significant improvement in accuracy and coverage in building full-length protein structures on various test sets compared to existing methods like RosettaES, MAINMAST, and Phenix.
Review
Pharmacology & Pharmacy
Yuyu Feng, Yumeng Yan, Jiahua He, Huanyu Tao, Qilong Wu, Sheng-You Huang
Summary: Nucleic acid-ligand interactions play crucial roles in cellular processes and have attracted significant interest in drug discovery. However, traditional docking algorithms and scoring functions for protein-ligand interactions may not be applicable to nucleic acid-ligand docking due to the differences in their properties. This review discusses the current status, challenges, and limitations of docking algorithms and scoring functions for DNA/RNA-ligand interactions.
DRUG DISCOVERY TODAY
(2022)
Article
Biochemical Research Methods
Linyuan Guo, Jiahua He, Peicong Lin, Sheng-You Huang, Jianxin Wang
Summary: This study introduces a new deep learning-based scoring method TRScore for ranking protein-protein docking models. TRScore effectively identifies near-native conformations in protein-protein docking models and shows significant improvement across various test sets.
Article
Multidisciplinary Sciences
Jiahua He, Peicong Lin, Ji Chen, Hong Cao, Sheng-You Huang
Summary: In this study, the authors propose an automatic model building method, named EMBuild, which can build high-quality complex structures from intermediate-resolution cryo-EM maps. By integrating multiple techniques and algorithms, EMBuild is able to accurately construct complex structures comparable to manually built ones.
NATURE COMMUNICATIONS
(2022)
Article
Multidisciplinary Sciences
Jiahua He, Tao Li, Sheng-You Huang
Summary: The authors propose a deep learning approach called EMReady to improve the quality and interpretability of cryo-EM maps by considering both local and non-local effects. Cryo-EM is crucial for determining the structure of macromolecular complexes, but raw cryo-EM maps often suffer from contrast loss and heterogeneity. EMReady is evaluated on diverse test sets and compared with state-of-the-art methods, demonstrating its ability to enhance map quality and improve model building.
NATURE COMMUNICATIONS
(2023)