Key result
FragBEST-Myo segments cardiac myosin omecamtiv mecarbil binding sites with ~95% accuracy.
Why the study?
Protein binding site features change with conformational dynamics, and mapping fragment-binding propensity onto binding site surfaces using deep learning could help monitor changes in ligand-binding capacity.
A novel deep learning tool accurately segments the omecamtiv mecarbil binding site on cardiac myosin, providing a proof-of-concept for using fragment-based mapping to improve ensemble docking and drug design.
Answers are grounded in this paper and cite their sources.
May aid computational refinement of myosin modulators; leaves open clinical translation given weak evidence for omecamtiv mecarbil.
The geometric and chemical features of protein binding sites tend to change as a consequence of conformational dynamics. In the ligand-unbound (apo) state, a binding site might be only transiently organized in a way that can accommodate a ligand, with the relevant regions of the protein coming together in a suitable arrangement only in a subset of conformations. Ligand binding itself can also induce changes in the site. Because most ligands can be decomposed into smaller fragments, we hypothesized that mapping onto the binding site surface the propensity of binding specific fragments could be used to monitor changes in the ability of the site to bind a ligand. This task can be formulated as semantic segmentation and addressed using deep learning. Here, we introduce the Fragment-Based protein Ensemble semantic Segmentation Tool for Myosin (FragBEST-Myo), a deep learning method based on a 3D U-Net architecture, trained to partition the omecamtiv mecarbil (OM) binding site of cardiac myosin into fragment-specific regions using only local shape and physicochemical features. The model was trained on labeled Molecular Dynamics trajectories of OM-bound myosin in both post-rigor and pre-powerstroke states, achieving an accuracy of ∼95% and a mean Intersection over Union (mIoU) > 0.76 on unseen trajectories from both states. When applied to apo trajectories, FragBEST-Myo-derived descriptors produced rankings consistent with similarity to holo conformations. Moreover, selecting apo frames based on FragBEST-Myo ranking increased the chance of recovering holo-like OM docking poses relative to randomly chosen control frames, supporting its use as a screening tool for ensemble docking. Beyond frame selection, fragment maps provide a compact representation to assess docking poses and to guide fragment-based design. Our proof of concept provides a basis for developing future general models applicable to a broader range of proteins and ligands, with the fragment-based formulation offering a natural route to generalization.
No takes yet. Share an insight, caveat, or question.
Yang et al. (2026) studied Cardiac myosin binding site prediction. FragBEST-Myo vs. Randomly chosen control frames was evaluated on Mean Intersection over Union (mIoU) and accuracy on unseen trajectories. FragBEST-Myo achieved an accuracy of ~95% and a mean Intersection over Union > 0.76 on unseen trajectories for segmenting the omecamtiv mecarbil binding site of cardiac myosin.
Synapse has enriched 10 related papers. Consider them for comparative context: