-
Efficient Generation of Membrane and Solvent Tetrahedral Meshes for Ion Channel Finite Element Calculation
Authors:
Zhen Chao,
Sheng Gui,
Benzhuo Lu,
Dexuan Xie
Abstract:
A finite element solution of an ion channel dielectric continuum model such as Poisson-Boltzmann equation (PBE) and a system of Poisson-Nernst-Planck equations (PNP) requires tetrahedral meshes for an ion channel protein region, a membrane region, and an ionic solvent region as well as an interface fitted irregular tetrahedral mesh of a simulation box domain. However, generating these meshes is ve…
▽ More
A finite element solution of an ion channel dielectric continuum model such as Poisson-Boltzmann equation (PBE) and a system of Poisson-Nernst-Planck equations (PNP) requires tetrahedral meshes for an ion channel protein region, a membrane region, and an ionic solvent region as well as an interface fitted irregular tetrahedral mesh of a simulation box domain. However, generating these meshes is very difficult and highly technical due to the related three regions having very complex geometrical shapes. Currently, an ion channel mesh generation software package developed in Lu's research group is one available in the public domain. To significantly improve its mesh quality and computer performance, in this paper, new numerical schemes for generating membrane and solvent meshes are presented and implemented in Python, resulting in a new ion channel mesh generation software package. Numerical results are then reported to demonstrate the efficiency of the new numerical schemes and the quality of meshes generated by the new package for ion channel proteins with ion channel pores having different geometric complexities.
△ Less
Submitted 14 January, 2022;
originally announced January 2022.
-
Molecular Sparse Representation by 3D Ellipsoid Radial Basis Function Neural Networks via $L_1$ Regularization
Authors:
Sheng Gui,
Zhaodi Chen,
Minxin Chen,
Benzhuo Lu
Abstract:
In this paper, we have developed an ellipsoid radial basis function neural network (ERBFNN) and algorithm for sparse representing of a molecular shape. To evaluate a sparse representation of the molecular shape model, the Gaussian density map of molecule is approximated by ERBFNN with a relatively small number of neurons. The deep learning models were trained by optimizing a nonlinear loss functio…
▽ More
In this paper, we have developed an ellipsoid radial basis function neural network (ERBFNN) and algorithm for sparse representing of a molecular shape. To evaluate a sparse representation of the molecular shape model, the Gaussian density map of molecule is approximated by ERBFNN with a relatively small number of neurons. The deep learning models were trained by optimizing a nonlinear loss function with $L_1$ regularization. Experimental results demonstrate that the original molecular shape is able to be represented with good accuracy by much fewer scale of ERBFNN by our algorithm. And our network in principle can be applied to multi-resolution sparse representation of molecular shape and coarse-grained molecular modeling.
△ Less
Submitted 9 May, 2020;
originally announced May 2020.
-
Sparse Representation of Gaussian Molecular Surface
Authors:
Sheng Gui,
Minxin Chen,
Benzhuo Lu
Abstract:
In this paper, we propose a model and algorithm for sparse representing Gaussian molecular surface. The original Gaussian molecular surface is approximated by a relatively small number of radial basis functions (RBFs) with rotational ellipsoid feature. The sparsity of the RBF representation is achieved by solving a nonlinear $L_1$ optimization problem. Experimental results demonstrate that the ori…
▽ More
In this paper, we propose a model and algorithm for sparse representing Gaussian molecular surface. The original Gaussian molecular surface is approximated by a relatively small number of radial basis functions (RBFs) with rotational ellipsoid feature. The sparsity of the RBF representation is achieved by solving a nonlinear $L_1$ optimization problem. Experimental results demonstrate that the original Gaussian molecular surface is able to be represented with good accuracy by much fewer RBFs using our $L_1$ model and algorithm. The sparse representation of Gaussian molecular surface is useful in various applications, such as molecular structure alignment, calculating molecular areas and volumes, and the method in principle can be applied to sparse representation of general shapes and coarse-grained molecular modeling.
△ Less
Submitted 19 September, 2019;
originally announced September 2019.
-
Non-uniform dependence for a generalized Degasperis-Procesi equation
Authors:
Shaohui Gui,
**lu Li,
Weipeng Zhu
Abstract:
In the paper, we consider the Cauchy problem for a generalized Degasperis-Procesi equation. We prove that the data-to-solution map is not uniformly continuous.
In the paper, we consider the Cauchy problem for a generalized Degasperis-Procesi equation. We prove that the data-to-solution map is not uniformly continuous.
△ Less
Submitted 6 March, 2018;
originally announced March 2018.
