Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2405.05864

SMARTINI3: Systematic Parametrization of Realistic Multi-Scale Membrane Models via Unsupervised Learning and Multi-Objective Evolutionary Algorithms

Alireza Soleimani·Herre Jelger Risselada
View original source ↗Read source document ↗Export metadata ↓

Abstract

In this study, we utilize genetic algorithms to develop a realistic implicit solvent ultra-coarse-grained (PC) membrane model comprising only three interaction sites. The key philosophy of the ultra-CG membrane model SMARTINI3 is its compatibility with realistic membrane proteins, for example, modeled within the Martini coarse-grained (CG) model, as well as with the widely used GROMACS software for molecular simulations. Our objective is to parameterize this ultra-CG model to accurately reproduce the experimentally observed structural and thermodynamic properties of PC membranes in real units, including properties such as area per lipid, area compressibility, bending modulus, line tension, phase transition temperature, density profile, and radial distribution function. In our example, we specifically focus on the properties of a POPC membrane, although the developed membrane model could be perceived as a generic model of lipid membranes. To optimize the performance of the model (the fitness), we conduct a series of evolutionary runs with diverse random initial population sizes (ranging from 96 to 384). We demonstrate that the ultra-CG membrane model we developed exhibits authentic lipid membrane behaviors, encompassing self-assembly into bilayers, vesicle formation, membrane fusion, and gel phase formation. Moreover, we demonstrate compatibility with the Martini coarse-grained model by successfully reproducing the behavior of a transmembrane domain embedded within a lipid bilayer. This facilitates the simulation of realistic membrane proteins within an ultra-CG bilayer membrane, enhancing the accuracy and applicability of our model in biophysical studies.

Explore related subjects

q-bio.BM

Keep this discovery

BibTeXRIS

Alireza Soleimani, Herre Jelger Risselada. 2024-05-09. SMARTINI3: Systematic Parametrization of Realistic Multi-Scale Membrane Models via Unsupervised Learning and Multi-Objective Evolutionary Algorithms. https://arxiv.org/abs/2405.05864

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Sequence-Informed Geometric Evaluation of RNA 3D Structures

Computational RNA structure pipelines generate many candidate conformations for the same sequence. Reliable evaluation therefore requires more than recognising plausible geometry, it requires determining whether that geometry is compatible with the sequence. We introduce SIRGE, a sequence-informed geometric evaluator that conditions structural representations on nucleotide embeddings from a pretrained RNA language model. Early results show that SIRGE outperforms established evaluators in Kendall--$\tau$ alignment, Top-1 selection, and Top-3 ranking. Controlled comparisons further show that sequence conditioning corrects errors made by an otherwise matched geometric model and improves target-level rank structure. These findings provide initial evidence that pretrained sequence representations supply ranking information that complements geometric reasoning.

q-bio.BM↗
arXiv2026

PocketVE: Stable and Property-Guided Structure-Based Drug Design with Variance-Exploding Diffusion

Protein-conditioned 3D molecule generation is a central challenge in structure-based drug design, requiring a balance between pocket compatibility, molecular properties, and physical geometry. We propose \textbf{PocketVE}, a protein-pocket-conditioned variance-exploding (VE) diffusion framework that couples stable coordinate denoising with inference-time property guidance. Specifically, PocketVE combines an EDM-style training and sampling setup for 3D denoising, classifier-free guidance for multi-property steering without external property classifiers, and adaptive protein perturbation as a training-time pocket regularizer. Evaluated on CrossDocked2020 under the GenBench3D protocol, PocketVE improves Valid$_{3\text{D}}$ from 58.6 to 80.6 and reduces strain energy from 457.4 to 127.9 relative to its TAGMol architectural baseline, while retaining competitive docking and molecular-property scores under moderate guidance. A guidance-scale study shows that moderate guidance gives a favorable balance between target-related objectives and geometric quality, whereas stronger guidance can degrade geometry and distributional fidelity. Pocket-permutation and PoseCheck diagnostics further support pocket-specific spatial compatibility with reduced steric conflicts. Overall, the results suggest that geometric stability and inference-time property guidance should be considered as coupled design objectives.

q-bio.BM↗
arXiv2026

Predicting directional flexibility in proteins

Predicting protein dynamics is a long-standing problem in computational structural biology. Often, protein function critically depends on local directed motions, such as hinge movements, catalytic loop rearrangements and domain reorientations, which can be characterized by directional flexibility and correlated structural motions of the protein backbone. While Molecular Dynamics (MD) simulations provide an established but often prohibitively expensive approach, recent deep generative models aim to reduce this cost by directly predicting conformational ensembles, emulating MD. However, due to their large size and the need to generate several states until the derived dynamical properties converge, these models remain expensive. In this work, we propose BackFlip-2: a fast SE(3)-equivariant graph neural network trained to directly predict dynamical descriptors, such as directional backbone flexibility and pairwise dynamic correlations, from an equilibrium structure. In a series of experiments, we show that our model matches the accuracy of substantially larger ensemble generation models while being orders of magnitude faster, and demonstrate that the proposed equivariant architecture is especially well-suited for capturing anisotropic motions in proteins. BackFlip-2 model weights, training and inference code are available at https://github.com/graeter-group/backflip.

q-bio.BM↗