INTEGRATIVE STRUCTURAL DYNAMICS: CONFORMATIONAL SAMPLING GUIDED BY SAXS AND CRYO-EM EXPERIMENTAL DATA

Conformational sampling of biomolecules remains a central challenge in computational biology, as classical molecular dynamics (MD) simulations frequently fail to capture large-amplitude motions essential to biological function. To address this limitation, we developed an integrative framework extending the Molecular Dynamics with Excited Normal Modes (MDeNM) method, which kinetically excites collective low-frequency modes to enhance conformational exploration. The framework incorporates heterogeneous experimental data from Small-Angle X-ray Scattering (SAXS) and Cryo-Electron Microscopy (Cryo-EM) to steer sampling toward structurally compatible regions of conformational space. Adenylate kinase (ADK) was used as a model system due to the availability of well-characterized open and closed experimental structures. Three protocols were compared: classical MD, MDeNM-SAXS, and MDeNM-SAXS+Cryo-EM. In integrative protocols, a stochastic selection criterion favors moves that improve agreement with experimental data, assessed via chi-square fit to SAXS curves and cross-correlation with Cryo-EM maps. Simulations were performed in NAMD with the CHARMM36m force field and Generalized Born implicit solvent. For each system, 40 replicas were generated varying re-excitation intervals (0.5-1 ps) and kinetic energies (20 and 50 kcal/mol). Performance was evaluated by RMSD relative to the target structure and structural quality metrics. Progressive integration of experimental restraints yielded systematic improvement in conformational convergence. The MDeNM-SAXS+Cryo-EM protocol achieved superior performance, reaching a mean RMSD of ~2 Å with low inter-replica dispersion within only 20 ps of simulation. In contrast, no structure sampled by classical MD fell below 6 Å RMSD. SAXS-only restraints produced intermediate results, and higher kinetic energy accelerated convergence at the cost of increased variability. Structural quality was preserved across all integrative solutions. These results demonstrate that experimental restraints substantially reduce the computational cost and timescale needed to sample large-amplitude conformational transitions, validating the integrative strategy and establishing Cryo-EM data as a key determinant for accurate structural fitting. Future work will explore the incorporation of additional data sources such as NMR and mass spectrometry, further expanding the applicability of this framework within integrative structural biology.