Mechanics and dynamics of Chignolin
Loading...
Date
2026-10-12
Authors
Γουσγουνέλης, Ανδρέας
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Chignolin is a designed ten-residue mini-protein (GYDPETGTWG) that forms a β-hairpin-like structure in water. Its small size does not make its conformational behavior trivial. Folded, partially folded and nonnative structures can all be sampled. This combination has made chignolin a useful model for moleculardynamics studies of protein folding. Here, nine cyclic mutations of its sequence are compared with wildtype (WT) chignolin to examine how residue order shapes the accessible conformational ensemble.
The simulations had been completed before the start of this thesis, the work reported here concerns trajectory processing and analysis. The initial comparison used Root-Mean-Square Deviation (RMSD), RootMean-Square Fluctuation (RMSF), radius of gyration (Rg) and Dictionary of Secondary Structure of Proteins (DSSP) to describe structural displacement, local mobility, compactness and secondary structure. Global tICA maps and system-specific free-energy landscapes then showed where WT and the mutants sampled the same structural space and where their distributions diverged.
A pooled MSM was explored as a diagnostic, but its implied timescales did not converge well enough to support a joint kinetic interpretation. Kinetic modeling was consequently confined to the eight WT trajectories. Their validated MSM contained three metastable states, with State 3 accounting for about 96% of the WT population. This dominant state had the lowest structural deviation, the greatest preservation of native contacts and the largest β-structure contribution. The rest of the states were rare alternatives. A balanced VAMPNet model supplied a nonlinear coordinate for cross-system comparison.
Mutant distance from WT was quantified with a structural reference built only from WT data. Frames inside its supported region were labelled WT-compatible and those outside it Novel. The Novel fraction varied widely, from roughly 25% to 74%. The frames that remained WT-compatible also occupied that region in different proportions from WT, as shown by bootstrap Jensen-Shannon distances.
The pooled Novel frames formed three reproducible structural basins rather than one diffuse non-WT ensemble. The response to the mutations was not a uniform loss of stability. Generally, residue order redistributed population among native-like conformations, WT-compatible alternatives and three recurrent non-WT families, including an expanded, contact-poor basin.
Description
Keywords
Chignolin, Protein folding, Cyclic sequence mutations