Emerging Themes in CryoEM─Single Particle Analysis Image Processing

Vilas, José Luis; Carazo, José-Marı́a; Sorzano, C.O.S.

doi:10.1021/acs.chemrev.1c00850

Cited by 24 publications

(10 citation statements)

References 229 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The recently developed method Xmipp Swarm consensus [ 98 ], available in Scipion [ 15 ], avoids user intervention by automatically calculating a consensus volume given a set of particles and initial maps obtained from different methods. This technique utilizes swarm optimization, a variation of SGD that incorporates momentum to accelerate convergence, to evolve the population of volumes towards a more globally correct solution [ 99 ].…”

Section: Conventional Approaches To Ab Initio Modeling and 3d Reconst...mentioning

confidence: 99%

Novel Artificial Intelligence-Based Approaches for Ab Initio Structure Determination and Atomic Model Building for Cryo-Electron Microscopy

DiIorio,

Kulczyk

2023

Micromachines

View full text Add to dashboard Cite

Single particle cryo-electron microscopy (cryo-EM) has emerged as the prevailing method for near-atomic structure determination, shedding light on the important molecular mechanisms of biological macromolecules. However, the inherent dynamics and structural variability of biological complexes coupled with the large number of experimental images generated by a cryo-EM experiment make data processing nontrivial. In particular, ab initio reconstruction and atomic model building remain major bottlenecks that demand substantial computational resources and manual intervention. Approaches utilizing recent innovations in artificial intelligence (AI) technology, particularly deep learning, have the potential to overcome the limitations that cannot be adequately addressed by traditional image processing approaches. Here, we review newly proposed AI-based methods for ab initio volume generation, heterogeneous 3D reconstruction, and atomic model building. We highlight the advancements made by the implementation of AI methods, as well as discuss remaining limitations and areas for future development.

show abstract

Section: Conventional Approaches To Ab Initio Modeling and 3d Reconst...mentioning

confidence: 99%

Novel Artificial Intelligence-Based Approaches for Ab Initio Structure Determination and Atomic Model Building for Cryo-Electron Microscopy

DiIorio,

Kulczyk

2023

Micromachines

View full text Add to dashboard Cite

show abstract

“…In other words, successive rounds of 2D and 3D classifications result in the discarding of many particles, and only those with consistent views and conformations are considered for final 3D reconstructions. 186 Therefore, the resulting maps provide a discrete view of the structural heterogeneity corresponding to highly populated free energy minima. Many of the methods presented in the previous sections enable the characterization of the transitions between those states and the identification of relevant intermediates.…”

Section: Dynamics and Heterogeneities Captured From The Images Or Ens...mentioning

confidence: 99%

“…Furthermore, information about low‐populated states is usually lost due to the application of image‐processing protocols required to obtain one or a few high‐resolution maps. In other words, successive rounds of 2D and 3D classifications result in the discarding of many particles, and only those with consistent views and conformations are considered for final 3D reconstructions 186 . Therefore, the resulting maps provide a discrete view of the structural heterogeneity corresponding to highly populated free energy minima.…”

Section: Dynamics and Heterogeneities Captured From The Images Or Ens...mentioning

confidence: 99%

Computational biophysics meets cryo‐EM revolution in the search for the functional dynamics of biomolecular systems

Costa,

Gur,

Krieger

et al. 2023

WIREs Comput Mol Sci

View full text Add to dashboard Cite

There is a variety of experimental and computational techniques available to explore protein dynamics, each presenting advantages and limitations. One promising experimental technique that is driving the development of computational methods is cryo‐electron microscopy (cryo‐EM). Cryo‐EM provides molecular‐level structural data and first estimates of conformational landscape from single particle analysis but cannot track real‐time protein dynamics and may contain uncertainties in atomic positions especially at highly dynamic regions. Molecular simulations offer atomic‐level insights into protein dynamics; however, their computing time requirements limit the conformational sampling accuracy, and it is often hard, to assess by full‐atomic simulations the cooperative movements of biological interest for large assemblies such as those resolved by cryo‐EM. Coarse‐grained (CG) simulations permit us to explore such systems, but at the costs of lower resolution and potentially incomplete sampling of conformational space. On the other hand, analytical methods may circumvent sampling limitations. In particular, elastic network models‐based normal mode analyses (ENM‐NMA) provide unique solutions for the complete mode spectra near equilibrium states, even for systems of megadaltons, and may thus deliver information on mechanisms of motions relevant to biological function. Yet, they lack atomic resolution as well as temporal information for non‐equilibrium systems. Given the complementary nature of these methods, the integration of molecular simulations and ENM‐NMA into hybrid methodologies has gained traction. This review presents the current state‐of‐the‐art in structure‐based computations and how they are helping us gain a deeper understanding of biological mechanisms, with emphasis on the development of hybrid methods accompanying the advances in cryo‐EM.This article is categorized under: Structure and Mechanism > Computational Biochemistry and Biophysics

show abstract

“…The new image may then be fed to a model as input, enabling the model to gain knowledge from a broader range of data to boost its accuracy. The demand for massive datasets can be alleviated by using image-wrapping augmentation to provide synthetic data for training models [31]. Figure 8 is an example of a wrapped image.…”

Section: Flippingmentioning

confidence: 99%

Efficient Detection of Hepatic Steatosis in Ultrasound Images Using Convolutional Neural Networks: A Comparative Study

Alshagathrh,

Musleh,

Alzubaidi

et al. 2023

View full text Add to dashboard Cite

Introduction: Artificial Intelligence (AI) is widely used in medical studies to interpret imaging data and improve the efficiency of healthcare professionals. Nonalcoholic fatty liver disease (NAFLD) is a common liver abnormality associated with an increased risk of hepatic cirrhosis, hepatocellular carcinoma, and cardiovascular morbidity and mortality. This study explores the use of AI for automated detection of hepatic steatosis in ultrasound images. Background: Ultrasound is a non-invasive, cost-effective, and widely available method for hepatic steatosis screening. However, its accuracy depends on the operator's expertise, necessitating automated methods to enhance diagnostic accuracy. AI, particularly Convolutional Neural Network (CNN) models, can provide accurate and efficient analysis of ultrasound images, enabling automated detection, improving diagnostic accuracy, and facilitating real-time analysis. Problem Statement: This study aims to evaluate deep learning methods for binary classification of hepatic steatosis using ultrasound images. Methodology: Open-source data is used to prepare three groups (A, B, C) of ultrasound images in different sizes. Images are augmented using seven pre-processing approaches (resizing, flipping, rotating, zooming, contrasting, brightening, and wrapping) to increase image variations. Seven CNN classifiers (EfficientNet-B0, ResNet34, AlexNet, DenseNet121, ResNet18, ResNet50, and MobileNet_v2) are evaluated using stratified 10fold cross-validation. Six metrics (accuracy, sensitivity, specificity, precision, F1 score, and MCC) are employed, and the best-performing fold epochs are selected. Experiments and Results: The study evaluates seven models, finding EfficientNet-B0, ResNet34, DenseNet121, and AlexNet to perform well in groups A and B. EfficientNet-B0 shows the best overall performance. It achieves high scores for all six metrics, with accuracy rates of 98.9%, 98.4%, and 96.3% in groups A, B, and C, respectively. Discussion and Conclusion: EfficientNet-B0, ResNet34, and DenseNet121 exhibit potential for classifying fatty liver ultrasound images. EfficientNet-B0 demonstrates the best average accuracy, specificity, and sensitivity, although more training data is needed for generalization. Complete and mediumsized images are preferred for classification. Further evaluation of other classifiers is necessary to determine the best model.

show abstract

Emerging Themes in CryoEM─Single Particle Analysis Image Processing

Cited by 24 publications

References 229 publications

Novel Artificial Intelligence-Based Approaches for Ab Initio Structure Determination and Atomic Model Building for Cryo-Electron Microscopy

Novel Artificial Intelligence-Based Approaches for Ab Initio Structure Determination and Atomic Model Building for Cryo-Electron Microscopy

Computational biophysics meets cryo‐EM revolution in the search for the functional dynamics of biomolecular systems

Efficient Detection of Hepatic Steatosis in Ultrasound Images Using Convolutional Neural Networks: A Comparative Study

Contact Info

Product

Resources

About