Simulation of lipid membrane models at coarse-grained resolution Towards novel methodologies with improved capabilities

This thesis develops computational methodologies for simulating biological membranes at coarse-grained (CG) resolution, combining well-established modeling strategies with recent advances in machine learning. It presents an optimized set of lipid representations and parameters for the Martini 3 CG force field (FF), improving the description of temperature-dependent phase separation in mixtures while also broadening the coverage of lipid types. Building on this foundation, a data-driven optimization protocol based on automatic differentiation is introduced, systematically refining CG FFs by minimizing a multi-objective loss function that combines bottom-up and top-down parametrization targets. To move beyond the intrinsic limitations of such models, a different approach is proposed, using graph neural networks trained on atomistic simulations to learn CG potential energy surfaces. Alongside these model-focused developments, this thesis presents tools to facilitate the analysis and dissemination of CG simulations: i- ProLint2, an efficient Python library for characterizing biomolecular interactions from molecular dynamics trajectories; and ii- a new web portal for the Martini Force Field Initiative, built on an automatically maintained infrastructure and backend operations that support open collaboration in developing and using CG models. Collectively, these contributions advance the methodological framework for simulating biological membranes at CG resolution, with significant applications to membrane biophysics.

Name thesis:degree_name

Doctor of Philosophy (PhD)

Discipline thesis:degree_discipline

Biological Sciences

Grantor

Science

Year dc:date.issued

2026

Author dc:creator

• Ramirez Echemendia, Daniel Pastor

Advisor dc:contributor.advisor

• Tieleman, D. Peter

Committee members dc:contributor.committeemember

• Prenner, Elmar
• Kusalik, Peter

• Biological membranes
• Molecular dynamics simulations
• Coarse-grained models
• Machine learning

Statement dc:rights

• Unless otherwise indicated, this material is protected by copyright and has been made available with authorization from the copyright owner. You may use this material in any way that is permitted by the Copyright Act or through licensing that has been assigned to the document. For uses that are not allowable under copyright legislation or licensing, you are required to seek permission.

Language dc:language.iso

en

DOI

https://dx.doi.org/10.11575/PRISM/51668

OAI identifier oai:identifier

oai:ucalgary.scholaris.ca:1880/125388