Biophysical Society Thematic Meeting | Hamburg 2022

Biophysics at the Dawn of Exascale Computers

Poster Abstracts

19-POS Board 19 TACKLING THE CONVERGENCE ISSUE OF N-GLYCAN CONFORMER DISTRIBUTIONS IN MOLECULAR DYNAMICS SIMULATION Isabell Louise Grothaus 1 ; Giovanni Bussi 2 ; Lucio Colombi Ciacchi 1 ; 1 University of Bremen, Hybrid Materials Interfaces Group, Faculty of Production Engineering, Bremen, Germany 2 Scuola Internazionale Superiore di Studi Avanzati (SISSA), Physics and Chemistry of Biological Systems, Trieste, Italy Branched polysaccharides, called N-glycans, represent the most diverse post-translational modification of proteins. Their fundamental role in proteins’ structure and function is becoming more recognized in molecular dynamics simulation only in recent years. N-glycans harbor an intrinsic flexibility due to their glycosidic linkages and monosaccharide ring dynamics, resulting in difficult-to-converge conformer populations in standard molecular dynamics simulation. We show that an enhanced-sampling metadynamics scheme based on enhancing transitions over all relevant barriers with concurrent one-dimensional energy potentials can in fact capture effectively all biologically relevant conformers of N-glycans. The framework can be applied to free and protein-bound glycans in solution, opening up new possibilities for the investigation of N-glycans in computational studies. Additionally, machine-learning based dimensionality reduction approaches are employed for an unambiguous representation of N-glycan conforms in a two-dimensional space. This visualization facilitates different kinds of analyses, such as force field or experimental parameter comparisons. The combination of presented techniques achieves converged N-glycan conformer populations in a reasonable time frame, dissecting the most varying and therefore important variables in the system, and hence allows to draw conclusions about force field accuracy or differences with respect to other N-glycan systems.

52