Exploring Local vs Global Models of SAR with Activity Atlas and Activity Miner
Rae Lawrence,
Technical Sales,
Cresset
The hundreds, if not thousands, of compounds prepared in lead optimization programs hold a wealth of information on potency, selectivity and ADMET properties, however, in many cases exploring the SAR information can prove daunting. To this end, we have recently presented two methods of navigating SAR landscapes using 3D activity cliff analyses.
Activity Atlas produces a global summary of the information obtained from 3D activity cliff analysis. Examination of all pairs of molecules can distinguish between apparent cliffs that are outliers, or due to measurement error, and those which consistently point to particular electrostatic and steric features having a large impact on activity. As part of this technique, a probabilistic map of the SAR of the series is developed. This provides an invaluable overview to the chemist, showing which parts of property space around a molecule have a strong effect on the activity.
Activity Miner is used to focus in on more specific activity cliffs, detected from 2D or 3D similarity metrics. Uniquely Activity Miner focusses on the reason for the cliff rather than simple detection. Using both 3D and 2D similarity metrics enables Activity Miner to detect more cliffs that either method detects in isolation. A small structural change could cause a large change in potency due to many different reasons, such as steric clash with the protein, the loss of (or gain of) hydrogen bond donors/acceptors, or through forcing an alternative conformation of the ligand. This technique automatically locates activity cliffs in the context of the target active site and clearly identifies the most important parts of the SAR, and visually shows the reason for the activity difference between each pair of compounds, assisting in a true understanding of the SAR landscape. Since large datasets may contain many activity cliffs it is important to use multiple view of the activity landscape that enable visualisation of the complete dataset or on the SAR around a specific compound.
In this talk, we present will introduce and demonstrate the utility of both methods as applied to several literature datasets to provide illumination for their structure-activity and structure-selectivity landscapes.
|
|
