The secreted glycoprotein Wnt family plays a critical role in cell development and numerous diseases, including cancer, are related to Wnt signaling malfunctions. Wnt proteins vary in sequence length and amino acid composition, but interestingly they all bind to the Wls transmembrane protein, that facilitates the transport to the cell membrane for secretion. Here, we combined extended atomistic molecular dynamics (MD) simulations and supervised machine learning (ML) method to unveil the molecular mechanisms underlying Wnt-Wls binding across the diverse Wnt family. We simulated four biologically significant Wnt-Wls complexes with multi-microsecond MD and developed a novel local structure alignment algorithm that enables systematic comparison of residue interactions across divergent Wnt sequences. After applying a two-stage clustering strategy to reduce feature redundancy and facilitate robust feature selection, we trained a Random Forest classifier that distinguished four Wnt classes with high accuracy. The feature analysis reveals both previously known and novel key residue pairs responsible for distinguishing among the Wnt systems, demonstrating the robustness of our approach and how interpretable ML can effectively uncover crucial biophysical interactions using MD trajectory data. Importantly, our integrated MD-ML framework provides a generalizable, data-driven approach for dissecting protein-protein interactions and guiding experimental validation or therapeutic targeting.
Callahan et al. (Sun,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: