Extended Data Fig. 1: Workflow to integrate X-ray structures, smFRET, bulk-FRET kinetics analysis, MD simulations and SAXS ensemble modelling.

Panels a–f summarize the smFRET workflow showing selection of smFRET constructs (a), determination of analysis methods (b), single-molecule data-pruning steps (c), maximum-information analysis (d), motional narrowing analysis (e) and smFRET constraints including peak positions of dye–dye distances and interconversion kinetic rates (f). For the trajectory shown in (d), the distance uncertainty was α = 0.11, which corresponded to an average time resolution of 〈∆〉 = 3.3 ms. For (e), at ∆=2.3 ms, the GrsA conformational PDF clearly displays two peaks. As the time resolution decreases (longer ∆), the two peaks coalesce into one at ∆ ≳5.1 ms. This motional-narrowing phenomenon shows that the timescales of GrsA A-___domain conformational dynamics overlap with those of smFRET experiment data acquisition. For (f1–f2), a quartile outlier test indicated that individual differences between smFRET measurements and bulk-FRET (f1) and those between smFRET measurements and unconstrained MD models (f2) are statistical. Panels g–h describe the bulk-FRET and smFRET integration and the associated kinetics modelling that determine the chemical state assignment, kinetic scheme, populations and rates. Panels i–j show the integration of smFRET and MD simulations. Panel k shows the smFRET and SAXS integration. One 10,000-structure pool was generated per A-___domain conformation and each point in the cloud plot represents the maximum distance from any atom in the PCP ___domain to the Cα of residue 527 at the C-terminus of the A^C-subdomain. The distance distribution is for the distance from the aforementioned PCP position to the NE atom of W239 located inside the Phe-binding pocket. Different combinations of pools were used to test the ability of various A-___domain conformations to allow for sufficient conformational sampling of the PCP ___domain to explain the experimental data.

Source data