An official website of the United States government
Abstract Cryo‐electron microscopy (cryo‐EM) has become a major experimental technique to determine the structures of large protein complexes and molecular assemblies, as evidenced by the 2017 Nobel Prize. Although cryo‐EM has been drastically improved to generate high‐resolution three‐dimensional maps that contain detailed structural information about macromolecules, the computational methods for using the data to automatically build structure models are lagging far behind. The traditional cryo‐EM model building approach is template‐based homology modeling. Manual de novo modeling is very time‐consuming when no template model is found in the database. In recent years, de novo cryo‐EM modeling using machine learning (ML) and deep learning (DL) has ranked among the top‐performing methods in macromolecular structure modeling. DL‐based de novo cryo‐EM modeling is an important application of artificial intelligence, with impressive results and great potential for the next generation of molecular biomedicine. Accordingly, we systematically review the representative ML/DL‐based de novo cryo‐EM modeling methods. Their significances are discussed from both practical and methodological viewpoints. We also briefly describe the background of cryo‐EM data processing workflow. Overall, this review provides an introductory guide to modern research on artificial intelligence for de novo molecular structure modeling and future directions in this emerging field. This article is categorized under:Structure and Mechanism > Molecular StructuresStructure and Mechanism > Computational Biochemistry and BiophysicsData Science > Artificial Intelligence/Machine Learning
more »
« less
Visualizing lipid membrane structure with cryo-EM: past, present, and future
The development of electron cryomicroscopy (cryo-EM) has evolved immensely in the last several decades and is now well-established in the analysis of protein structure both in isolation and in their cellular context. This review focuses on the history and application of cryo-EM to the analysis of membrane architecture. Parallels between the levels of organization of protein structure are useful in organizing the discussion of the unique parameters that influence membrane structure and function. Importantly, the timescales of lipid motion in bilayers with respect to the timescales of sample vitrification is discussed and reveals what types of membrane structure can be reliably extracted in cryo-EM images of vitrified samples. Appreciating these limitations, a review of the application of cryo-EM to examine the lateral organization of ordered and disordered domains in reconstituted and biologically derived membranes is provided. Finally, a brief outlook for further development and application of cryo-EM to the analysis of membrane architecture is provided.
more »
« less
- PAR ID:
- 10424242
- Publisher / Repository:
- Portland Press
- Date Published:
- Journal Name:
- Emerging Topics in Life Sciences
- Volume:
- 7
- Issue:
- 1
- ISSN:
- 2397-8554
- Page Range / eLocation ID:
- 55 to 65
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
null (Ed.)Rotary vacuolar adenosine triphosphatases (V-ATPases) drive transmembrane proton transport through a V o proton channel subcomplex. Despite recent high-resolution structures of several rotary ATPases, the dynamic mechanism of proton pumping remains elusive. Here, we determined a 2.7-Å cryo–electron microscopy (cryo-EM) structure of yeast V o proton channel in nanodisc that reveals the location of ordered water molecules along the proton path, details of specific protein-lipid interactions, and the architecture of the membrane scaffold protein. Moreover, we uncover a state of V o that shows the c -ring rotated by ~14°. Molecular dynamics simulations demonstrate that the two rotary states are in thermal equilibrium and depict how the protonation state of essential glutamic acid residues couples water-mediated proton transfer with c -ring rotation. Our cryo-EM models and simulations also rationalize a mechanism for inhibition of passive proton transport as observed for free V o that is generated as a result of V-ATPase regulation by reversible disassembly in vivo.more » « less
-
The nanoscale organization of biological membranes into structurally and compositionally distinct lateral domains is believed to be central to membrane function. The nature of this organization has remained elusive due to a lack of methods to directly probe nanoscopic membrane features. We show here that cryogenic electron microscopy (cryo-EM) can be used to directly image coexisting nanoscopic domains in synthetic and bioderived membranes without extrinsic probes. Analyzing a series of single-component liposomes composed of synthetic lipids of varying chain lengths, we demonstrate that cryo-EM can distinguish bilayer thickness differences as small as 0.5 Å, comparable to the resolution of small-angle scattering methods. Simulated images from computational models reveal that features in cryo-EM images result from a complex interplay between the atomic distribution normal to the plane of the bilayer and imaging parameters. Simulations of phase-separated bilayers were used to predict two sources of contrast between coexisting ordered and disordered phases within a single liposome, namely differences in membrane thickness and molecular density. We observe both sources of contrast in biomimetic membranes composed of saturated lipids, unsaturated lipids, and cholesterol. When extended to isolated mammalian plasma membranes, cryo-EM reveals similar nanoscale lateral heterogeneities. The methods reported here for direct, probe-free imaging of nanodomains in unperturbed membranes open new avenues for investigation of nanoscopic membrane organization.more » « less
-
Elucidating protein–ligand interaction is crucial for studying the function of proteins and compounds in an organism and critical for drug discovery and design. The problem of protein–ligand interaction is traditionally tackled by molecular docking and simulation, which is based on physical forces and statistical potentials and cannot effectively leverage cryo-EM data and existing protein structural information in the protein–ligand modeling process. In this work, we developed a deep learning bioinformatics pipeline (DeepProLigand) to predict protein–ligand interactions from cryo-EM density maps of proteins and ligands. DeepProLigand first uses a deep learning method to predict the structure of proteins from cryo-EM maps, which is averaged with a reference (template) structure of the proteins to produce a combined structure to add ligands. The ligands are then identified and added into the structure to generate a protein–ligand complex structure, which is further refined. The method based on the deep learning prediction and template-based modeling was blindly tested in the 2021 EMDataResource Ligand Challenge and was ranked first in fitting ligands to cryo-EM density maps. These results demonstrate that the deep learning bioinformatics approach is a promising direction for modeling protein–ligand interactions on cryo-EM data using prior structural information.more » « less
-
method for structure determination. Despite the substantial growth in deposited cryo-EM maps driven by advances in microscopy and image processing, accurately constructing models from these maps remains challenging. Extracting secondary structure information from EM maps is valuable for cryo-EM modeling. In this context, we introduce a novel deep learning secondary structure annotation framework specifically designed for intermediate-resolution cryo-EM maps, employing a three-dimensional Inception architecture. Testing it on diverse datasets, including maps with authentic intermediate resolutions, demonstrates its accuracy and robustness in identifying secondary structures in cryo-EM maps. We conducted a comparative analysis of our results against frameworks that exist in the state-of-the-art, and our framework demonstrated superior performance across nearly all secondary structure elements. We employed the F1 accuracy metric, yielding an average F1 score of 0.657 for helix, 0.712 for coil, and 0.596 for sheet predictions. Notably, certain helix and sheet predictions achieved an impressive F1 score of 0.881.more » « less
