Abstract:Abstract An increasing number of density maps of biological macromolecules have been determined by cryo-electron microscopy (cryo-EM) and stored in the public database, EMDB. To interpret the structural information contained in EM density maps, alignment of maps is an essential step for structure modeling, comparison of maps, and for database search. Here, we developed VESPER, which captures the similarity of underlying molecular structures embedded in density maps by taking local gradient directions into consideration. Compared to existing methods, VESPER achieved substantially more accurate global and local alignment of maps as well as database retrieval.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is to improve the global and local alignment accuracy of cryo - EM (cryo - electron microscopy) density maps and enhance the accuracy of database retrieval. Specifically: 1. **Background problems**: - With the development of cryo - electron microscopy technology, more and more biomacromolecular structures are determined by cryo - electron microscopy and stored in public databases (such as EMDB). - In order to interpret the structural information in these density maps, the alignment of density maps is a crucial step. Alignment operations are essential for structural modeling, map comparison, and database searches. 2. **Limitations of existing methods**: - Existing methods (such as gm fit and fitmap) have limitations in global and local alignment, especially performing poorly in partial matching. - These methods also show limited accuracy when dealing with the sub - structure fitting of complex structures. 3. **VESPER's goals**: - VESPER aims to improve the global and local alignment accuracy of cryo - EM density maps by introducing a new alignment method. - VESPER uses local density vectors to capture the molecular structure information embedded in the density maps and evaluates the alignment effect by calculating the dot product of the matching vectors. - Compared with existing methods, VESPER shows higher alignment accuracy and database retrieval ability on the benchmark data set. ### Working principle of VESPER The core idea of VESPER is to represent each density map as a set of unit vectors pointing to neighboring high - density points. The directions of these vectors capture the local structural information embedded in the density map. The specific steps are as follows: 1. **Vector representation**: - Calculate the unit vectors of each density map using the mean - shift algorithm. - The vector representation captures the local molecular structure information around each voxel. 2. **Alignment scoring**: - Evaluate the alignment effect by summing the dot products of the matching vectors in the two aligned density maps. - Use the fast Fourier transform (FFT) for rotation and translation searches to find the optimal alignment position. 3. **Optimization and refinement**: - For the top 10 highest - scoring models, perform a more refined angular search (default 5° interval). - Output the highest - scoring alignment result. ### Experimental results VESPER has demonstrated its superior performance in multiple experiments, especially in global and local alignment and database retrieval. Experimental results show that VESPER exhibits high accuracy and robustness in density map alignment within different resolution ranges. In conclusion, the main goal of this paper is to develop a new method - VESPER - to improve the global and local alignment accuracy of cryo - EM density maps and improve the accuracy of database retrieval.

VESPER: global and local cryo-EM map alignment using local density vectors

Accurate global and local 3D alignment of cryo-EM density maps using local spatial structural features

CryoAlign: feature-based method for global and local 3D alignment of EM density maps

Protein complex structure modeling by cross-modal alignment between cryo-EM maps and protein sequences

Improvement of cryo-EM maps by density modification

CryoRes: Local Resolution Estimation of Cryo-EM Density Maps by Deep Learning.

Local computational methods to improve the interpretability and analysis of cryo-EM maps

Enhancing Density Maps by Removing the Majority of Particles in Single Particle Cryogenic Electron Microscopy Final Stacks

Model-based local density sharpening of cryo-EM maps

Improvement of cryo-EM maps by simultaneous local and non-local deep learning

Recognizing amino acid sidechains in a medium resolution cryo-electron density map

Residue-wise local quality estimation for protein models from cryo-EM maps

EMPOT: partial alignment of density maps and rigid body fitting using unbalanced Gromov-Wasserstein divergence

Auto3DCryoMap: an automated particle alignment approach for 3D cryo-EM density map reconstruction

De Novo Protein Structure Determination from Near-Atomic-resolution Cryo-Em Maps.

Accurate Model Annotation of A Near-Atomic Resolution Cryo-Em Map

EMDA: A Python package for Electron Microscopy Data Analysis

Atomic Accuracy Models from 4.5 Å Cryo-Electron Microscopy Data with Density-Guided Iterative Local Refinement

Cryo-EM structure and B-factor refinement with ensemble representation

CryoFM: A Flow-based Foundation Model for Cryo-EM Densities

De Novo Atomic Protein Structure Modeling for Cryo-EM Density Maps Using 3D Transformer and Hidden Markov Model