Quantitative predictions based on physically realistic and interpretable models
Surflex-QMOD addresses the physical linkage between a model for predicting ligand affinity and molecular binding modes. As a result, Surflex-QMOD provides more than a mathematical model for numerical predictions of binding affinities but also provides a physically interpretable model of the protein binding pocket and ligand poses that explain the physical properties responsible for ligand binding. By design, this enables one to make a direct correspondence between the physical process of protein-ligand binding and the act of prediction.
With Surflex-QMOD, discovery scientists can:
- Make quantitative predictions of a drug candidate's affinity for its target
- Assess the confidence of quantitative predictions
- Quantify the novelty of drug candidate molecules
- Leverage SAR from one chemotype to make accurate predictions for novel chemotypes.
- Combine SAR information from multiple chemotypes into a single model.
Using ligand SAR data, Surflex-QMOD constructs a physically realistic model of a binding pocket and the biologically relevant poses of the ligands. Although protein structural information is not required for constructing accurate Surflex-QMOD models, this data can be combined with ligand SAR data to improve both model accuracy and applicability. Surflex-QMOD provides a powerful bridge that allows you to integrate ligand SAR information and receptor structure information to generate physically consistent and highly predictive models. Surflex-QMOD has demonstrated predictive accuracy on a broad domain of therapeutically relevant targets such on enzymes, G-protein coupled receptors, and ligand-gated ion channels.