Modelling Complex Structures from Limited Crosslinks

We assessed IMPROViSeD's ability to accurately model complex structures using a limited number of crosslinks, aiming to evaluate its robustness in the face of incomplete or ambiguous data, which is common in experimental settings.

Predicting New Orientations of Subunits

We tested the ability of the algorithm to model new orientations of subunits within a complex, distinct from those available in the PDB repository, while still adhering to crosslinking constraints. This evaluated the method's capacity to predict novel interfaces and conformations that are absent from current experimental datasets.

We further explored the potential of IMPROViSeD to model structures for complexes using both experimental crosslinking data and synthetic data, demonstrating its flexibility and applicability across different data sources to achieve a consistent and accurate structural representation.