Molecular Surface Remeshing with Local Region Refinement

Dawar Khan, Dong‐Ming Yan, Sheng Gui, Benzhuo Lu, Xiaopeng ZhangInternational Journal of Molecular Sciences 2018 · DOI

TL;DR

This paper introduces a novel mesh refinement technique for molecular surfaces, crucial for accurate implicit-solvent modeling in computational biology. The method effectively smooths meshes, removes problematic regions with small angles and redundant vertices, and significantly improves overall mesh quality, outperforming existing state-of-the-art approaches.

What Problem Does This Paper Address?

Generating high-quality surface meshes for molecules is a critical first step for various computational biology applications, particularly for implicit-solvent modeling using methods like the Boundary Element Method (BEM) and Finite Element Method (FEM). Raw molecular surface meshes, often generated by tools like TMSmesh, are plagued by issues such as small angles, redundant vertices, and generally low-quality elements. These defects hinder downstream simulations and analyses. While existing remeshing methods exist, they often struggle to efficiently address these specific challenges in complex molecular geometries, leading to a demand for more robust and effective refinement techniques.

Methods

The proposed method, "Molecular Surface Remeshing with Local Region Refinement," employs a multi-stage approach:

  1. Initialization with Real-Time Adaptive Remeshing (RAR): The process begins by applying the RAR method to the initial mesh. RAR is chosen for its relative simplicity, efficiency, and ability to adaptively size mesh elements based on local curvature and error tolerance.
  2. Aspect Ratio Improvement: Following RAR, a global smoothing method is used to enhance the aspect ratios of the mesh triangles. This step involves calculating new vertex positions based on Voronoi cell centers and Laplacian smoothing, translating vertices only if the aspect ratio improves.
  3. Cut-and-Fill Strategy: This is the core novel component. It addresses problematic regions with small angles (<15° or <30°) and redundant vertices.
  4. Cut: Invalid regions (triangles with small angles) are identified and "cut" out.
  5. Fill: The resulting holes are then "filled" by connecting boundary edges. For larger holes, new vertices are inserted.
  6. Local Smooth: The newly filled regions are then locally smoothed using a combination of edge collapse (for short edges) and vertex translation (based on Laplacian centers) to improve minimal angles. This cut-and-fill-smooth cycle is applied twice, first with a 15° threshold and then with a 30° threshold.
  7. Surface Repair: Finally, a surface repair module is used to eliminate any remaining self-intersections or non-manifold elements, ensuring a clean mesh for subsequent tetrahedralization.

Datasets

The study utilized molecular surface meshes derived from structures found in the Protein Data Bank (PDB). Corresponding PQR files were generated using the pdb2pqr software. The benchmark dataset included several molecular structures: 1MAG, 2JK4, 1bl8, NaR1R4, AChE, and Connexin. These structures were used for both generating input meshes and evaluating the performance of the proposed method against existing ones.

Key Insights and Ideas

  • Targeted Defect Removal: The paper highlights that simply smoothing meshes isn't enough for molecular surfaces. A specific "cut-and-fill" strategy is crucial for effectively removing problematic small-angle triangles and redundant vertices that plague raw molecular meshes.
  • Hierarchical Refinement: Applying the cut-and-fill strategy with progressively larger angle thresholds (15° then 30°) allows for a more controlled and thorough refinement process, addressing both very small and moderately small angles.
  • Local Operators for Global Improvement: The method effectively combines global aspect ratio improvement with localized cut-and-fill and smoothing operations, demonstrating a balanced approach to mesh quality enhancement.
  • Importance of Mesh Quality for Simulations: The study underscores that poor mesh quality directly impacts the success of downstream simulations, such as Poisson-Boltzmann electrostatics calculations, where low-quality meshes can lead to failed computations.

Key Results

  • Significant Improvement in Minimal Angle: The proposed method achieved a minimal angle of at least 30° for all tested molecular surfaces, a substantial improvement over methods like RAR which left angles smaller than 30°.
  • Superior Performance: Compared to four state-of-the-art methods (ISO2mesh, Taubin method, Graphite, and SMOPT), the proposed method demonstrated significant improvements in minimal angle, aspect ratio, and other meshing quality metrics.
  • Area and Volume Preservation: The method showed satisfactory results in preserving the area and volume of the original molecular surfaces, with only minor deviations compared to other methods.
  • TetGen and AFMPB Compatibility: The refined meshes generated by the proposed method consistently passed TetGen for tetrahedral mesh generation and produced convergent results for AFMPB (Poisson-Boltzmann solver), unlike some other methods that failed.
  • High Percentage of Regular Vertices: The method achieved a high percentage of regular vertices, indicating a more uniform and well-structured mesh.

Limitations and Future Work

  • Area/Volume Preservation: While satisfactory, the paper acknowledges that the Taubin method and Graphite achieved slightly superior area/volume preservation. The proposed method performs intermediately in this regard.
  • Efficiency: The paper mentions plans to improve the efficiency of the method and develop it into a software product.
  • Mesh Generation from PQR: Future work includes introducing a mesh generation method for extracting surface meshes directly from PQR files with minimal defects.
  • Tetrahedral Generation: The authors plan to contribute to tetrahedral generation for molecular surface meshes.

Summary updated

Abstract

Molecular surface mesh generation is a prerequisite for using the boundary element method (BEM) and finite element method (FEM) in implicit-solvent modeling. Molecular surface meshes typically have small angles, redundant vertices, and low-quality elements. In the implicit-solvent modeling of biomolecular systems it is usually required to improve the mesh quality and eliminate low-quality elements. Existing methods often fail to efficiently remove low-quality elements, especially in complex molecular meshes. In this paper, we propose a mesh refinement method that smooths the meshes, eliminates invalid regions in a cut-and-fill strategy, and improves the minimal angle. We compared our method with four different state-of-the-art methods and found that our method showed a significant improvement over state-of-the-art methods in minimal angle, aspect ratio, and other meshing quality measurements. In addition, our method showed satisfactory results in terms of the ratio of regular vertices and the preservation of area and volume.
BibTeX
@article{khan2018molecular,
  author = {Dawar Khan and Dong‐Ming Yan and Sheng Gui and Benzhuo Lu and Xiaopeng Zhang},
  title = {{Molecular Surface Remeshing with Local Region Refinement}},
  journal = {International Journal of Molecular Sciences},
  year = {2018},
  doi = {10.3390/ijms19051383},
}
Figure