Discover and optimize new lead compounds using quantitative predictions of binding-site chemistry
The Advantages of Field-Based QSAR
Quantitative structure-activity relationships (QSAR) have long been a favorite way for researchers to optimize lead compounds. However, traditional QSAR models typically incorporate only crude approximations of 3D structure.
Field-Based QSAR disposes of this limitation. Beginning with a set of aligned ligands that have known activities, Field-Based QSAR is capable of inferring how the ligand’s electrostatic, hydrophobic, and steric fields result in biological activity or inactivity.
Like all ligand-based approaches, Field-Based QSAR requires no knowledge of receptor structure. Ideal for both lead discovery and lead optimization, Field-Based QSAR is capable of quickly turning existing data sets into useful QSAR models, helping researchers to leapfrog around patent space, synthetic roadblocks, and ADME restrictions.
Exercise fine-grained control over model-building:
Field-Based QSAR gives users a fine degree of control over model building. Working from the graphical interface, researchers can adjust the most important model-building parameters. Specialists will find a plethora of additional options available to them on the command line.
Look beyond R2 values:
Field-Based QSAR makes it easy to validate the predictive power of your model. Supplied with ligands that have known activities, Field-Based QSAR can randomly split them into training and test sets. Before synthesizing and assaying compounds on the basis of a QSAR prediction, researchers can evaluate the quality of their test set predictions in order to vet QSAR models for overfitting and other problems. In addition to correlation coefficients, Field-Based QSAR reports all widely used statistical metrics and offers 2D plotting tools.
View 3D volumes:
In addition to its obvious usefulness as a quantitative tool, Field-Based QSAR also helps researchers define the general, qualitative characteristics of active ligands. QSAR contour visualization shows where (and where not) to place steric volume, hydrophobic groups, and positively or negatively charged functional groups.
Citations and Acknowledgements
Schrödinger Release 2017-1: Field-based QSAR, Schrödinger, LLC, New York, NY, 2017.