A Tool for Segmentation of Secondary Structures in 3D Cryo-EM Density Map Components Using Deep Convolutional Neural Networks

Visualizing the virus world inside the cell by cryo-electron tomography

Structural studies on purified virus have revealed intricate architectures, but there is little structural information on how viruses interact with host cells in situ. Cryo-focused ion beam (FIB) milling and cryo-electron tomography (cryo-ET) have emerged as revolutionary tools in structural biology to visualize the dynamic conformational of viral particles and their interactions with host factors within infected cells. Here, we review the state-of-the-art cryo-ET technique for in situ viral structure studies and highlight exemplary studies that showcase the remarkable capabilities of cryo-ET in capturing the dynamic virus-host interaction, advancing our understanding of viral infection and pathogenesis.

The combined focal loss and dice loss function improves the segmentation of beta-sheets in medium-resolution cryo-electron-microscopy density maps

Summary

Although multiple neural networks have been proposed for detecting secondary structures from medium-resolution (5-10 Å) cryo-electron microscopy (cryo-EM) maps, the loss functions used in the existing deep learning networks are primarily based on cross-entropy loss, which is known to be sensitive to class imbalances. We investigated five loss functions: cross-entropy, Focal loss, Dice loss, and two combined loss functions. Using a U-Net architecture in our DeepSSETracer method and a dataset composed of 1355 box-cropped atomic-structure/density-map pairs, we found that a newly designed loss function that combines Focal loss and Dice loss provides the best overall detection accuracy for secondary structures. For β-sheet voxels, which are generally much harder to detect than helix voxels, the combined loss function achieved a significant improvement (an 8.8% increase in the F1 score) compared to the cross-entropy loss function and a noticeable improvement from the Dice loss function. This study demonstrates the potential for designing more effective loss functions for hard cases in the segmentation of secondary structures. The newly trained model was incorporated into DeepSSETracer 1.1 for the segmentation of protein secondary structures in medium-resolution cryo-EM map components. DeepSSETracer can be integrated into ChimeraX, a popular molecular visualization software.

Availability and implementation

https://www.cs.odu.edu/∼bioinfo/B2I_Tools/.

Automated model-free analysis of cryo-EM volume ensembles with SIREn

Cryogenic electron microscopy (cryo-EM) has the potential to capture snapshots of proteins in motion and generate hypotheses linking conformational states to biological function. This potential has been increasingly realized by the advent of machine learning models that allow 100s-1,000s of 3D density maps to be generated from a single dataset. How to identify distinct structural states within these volume ensembles and quantify their relative occupancies remain open questions. Here, we present an approach to inferring variable regions directly from a volume ensemble based on the statistical co-occupancy of voxels, as well as a 3D-convolutional neural network that predicts binarization thresholds for volumes in an unbiased and automated manner. We show that these tools recapitulate known heterogeneity in a variety of simulated and real cryo-EM datasets, and highlight how integrating these tools with existing data processing pipelines enables improved particle curation and the construction of quantitative conformational landscapes.

Determining Protein Secondary Structures in Heterogeneous Medium-Resolution Cryo-EM Images Using CryoSSESeg

While the acquisition of cryo-electron microscopy (cryo-EM) at near-atomic resolution is becoming more prevalent, a considerable number of density maps are still resolved only at intermediate resolutions (5-10 Å). Due to the large variation in quality among these medium-resolution density maps, extracting structural information from them remains a challenging task. This study introduces a convolutional neural network (CNN)-based framework, cryoSSESeg, to determine the organization of protein secondary structure elements in medium-resolution cryo-EM images. CryoSSESeg is trained on approximately 1300 protein chains derived from around 500 experimental cryo-EM density maps of varied quality. It demonstrates strong performance with residue-level F 1 scores of 0.76 for helix detection and 0.60 for β-sheet detection on average across a set of testing chains. In comparison to traditional image processing tools like SSETracer, which demand significant manual intervention and preprocessing steps, cryoSSESeg demonstrates comparable or superior performance. Additionally, it demonstrates competitive performance alongside another deep learning-based model, Emap2sec. Furthermore, this study underscores the importance of secondary structure quality, particularly adherence to expected shapes, in detection performance, emphasizing the necessity for careful evaluation of the data quality.

Untangling Irregular Actin Cytoskeleton Architectures in Tomograms of the Cell with Struwwel Tracer

In this work, we established, validated, and optimized a novel computational framework for tracing arbitrarily oriented actin filaments in cryo-electron tomography maps. Our approach was designed for highly complex intracellular architectures in which a long-range cytoskeleton network extends throughout the cell bodies and protrusions. The irregular organization of the actin network, as well as cryo-electron-tomography-specific noise, missing wedge artifacts, and map dimensions call for a specialized implementation that is both robust and efficient. Our proposed solution, Struwwel Tracer, accumulates densities along paths of a specific length in various directions, starting from locally determined seed points. The highest-density paths originating from the seed points form short linear candidate filament segments, which are further scrutinized and classified by users via inspection of a novel pruning map, which visualizes the likelihood of being a part of longer filaments. The pruned linear candidate filament segments are then iteratively fused into continuous, longer, and curved filaments based on their relative orientations, gap spacings, and extendibility. When applied to the simulated phantom tomograms of a Dictyostelium discoideum filopodium under experimental conditions, Struwwel Tracer demonstrated high efficacy, with F1-scores ranging from 0.85 to 0.90, depending on the noise level. Furthermore, when applied to a previously untraced experimental tomogram of mouse fibroblast lamellipodia, the filaments predicted by Struwwel Tracer exhibited a good visual agreement with the experimental map. The Struwwel Tracer framework is highly time efficient and can complete the tracing process in just a few minutes. The source code is publicly available with version 3.2 of the free and open-source Situs software package.

Spaghetti Tracer: A Framework for Tracing Semiregular Filamentous Densities in 3D Tomograms

Within cells, cytoskeletal filaments are often arranged into loosely aligned bundles. These fibrous bundles are dense enough to exhibit a certain regularity and mean direction, however, their packing is not sufficient to impose a symmetry between-or specific shape on-individual filaments. This intermediate regularity is computationally difficult to handle because individual filaments have a certain directional freedom, however, the filament densities are not well segmented from each other (especially in the presence of noise, such as in cryo-electron tomography). In this paper, we develop a dynamic programming-based framework, Spaghetti Tracer, to characterizing the structural arrangement of filaments in the challenging 3D maps of subcellular components. Assuming that the tomogram can be rotated such that the filaments are oriented in a mean direction, the proposed framework first identifies local seed points for candidate filament segments, which are then grown from the seeds using a dynamic programming algorithm. We validate various algorithmic variations of our framework on simulated tomograms that closely mimic the noise and appearance of experimental maps. As we know the ground truth in the simulated tomograms, the statistical analysis consisting of precision, recall, and F1 scores allows us to optimize the performance of this new approach. We find that a bipyramidal accumulation scheme for path density is superior to straight-line accumulation. In addition, the multiplication of forward and backward path densities provides for an efficient filter that lifts the filament density above the noise level. Resulting from our tests is a robust method that can be expected to perform well (F1 scores 0.86-0.95) under experimental noise conditions.

Tracing Filaments in Simulated 3D Cryo-Electron Tomography Maps Using a Fast Dynamic Programming Algorithm

We propose a fast, dynamic programming-based framework for tracing actin filaments in 3D maps of subcellular components in cryo-electron tomography. The approach can identify high-density filament segments in various orientations, but it takes advantage of the arrangement of actin filaments within cells into more or less tightly aligned bundles. Assuming that the tomogram can be rotated such that the filaments can be oriented to be directed in a dominant direction (i.e., the X, Y, or Z axis), the proposed framework first identifies local seed points that form the origin of candidate filament segments (CFSs), which are then grown from the seeds using a fast dynamic programming algorithm. The CFS length l can be tuned to the nominal resolution of the tomogram or the separation of desired features, or it can be used to restrict the curvature of filaments that deviate from the overall bundle direction. In subsequent steps, the CFSs are filtered based on backward tracing and path density analysis. Finally, neighboring CFSs are fused based on a collinearity criterion to bridge any noise artifacts in the 3D map that would otherwise fractionalize the tracing. We validate our proposed framework on simulated tomograms that closely mimic the features and appearance of experimental maps.

Combining Cryo-EM Density Map and Residue Contact for Protein Secondary Structure Topologies

Although atomic structures have been determined directly from cryo-EM density maps with high resolutions, current structure determination methods for medium resolution (5 to 10 Å) cryo-EM maps are limited by the availability of structure templates. Secondary structure traces are lines detected from a cryo-EM density map for α-helices and β-strands of a protein. A topology of secondary structures defines the mapping between a set of sequence segments and a set of traces of secondary structures in three-dimensional space. In order to enhance accuracy in ranking secondary structure topologies, we explored a method that combines three sources of information: a set of sequence segments in 1D, a set of amino acid contact pairs in 2D, and a set of traces in 3D at the secondary structure level. A test of fourteen cases shows that the accuracy of predicted secondary structures is critical for deriving topologies. The use of significant long-range contact pairs is most effective at enriching the rank of the maximum-match topology for proteins with a large number of secondary structures, if the secondary structure prediction is fairly accurate. It was observed that the enrichment depends on the quality of initial topology candidates in this approach. We provide detailed analysis in various cases to show the potential and challenge when combining three sources of information.