FGFR1 resistance mutations alter erdafitinib-binding dynamics, not binding energy, in simulations
- Method: all-atom MD of erdafitinib-bound FGFR1 kinase domain, wild type plus gatekeeper mutant V561M and molecular-brake mutant N546K, three replicates each, scored with MM/GBSA and MM/PBSA.
- Neither end-point method detected a mutation-associated change in binding free energy; differences fell within method uncertainty.
- N546K sharply cut occupancy and dwell time of the residue-546–Glu562 contact while the standard Cα–Cα brake criterion still read intact; V561M kept the brake but loosened the gatekeeper-flanking hydrophobic network.
- Validation is simulation-only, with no experimental data.
💡 Points
A null from MM/GBSA and MM/PBSA is expected here: end-point methods are too noisy to resolve small mutant-versus-wild-type shifts, so it isn't evidence of equal affinity. The real finding is that the Cα–Cα brake criterion missed N546K's lost Glu562 contact; check your own brake readouts. Still, three replicates, no experimental IC50 or resistance data, and no free-energy perturbation leave this hypothesis-generating.