Knowledge Base

Article ID: 125 - Last Modified: December 4, 2010

Should excluded volumes be used in the Phase QSAR model building process?

It is not recommend to incorporate excluded volumes into the QSAR model creation process. This is because the QSAR model itself already has the capacity to encode sterically forbidden regions. For example, if some bulky group appears at a particular location only within the some subset of the inactives, Phase QSAR may very well incorporate that information into the model by penalizing ligands that occupy this region. But this only works when the ligands are allowed to align in the most consistent manner, so that the favorable parts of structure are superimposed and the parts that would clash with the receptor get stuck into sterically forbidden regions, i.e., into the hypothetical receptor wall.

If you place excluded volumes in the sterically forbidden regions and realign the inactives, then one of two things will happen for a given inactive: (1) it will be impossible to find an alignment that matches the hypothesis on at least 3 points and avoids the excluded volumes; or (2) some alignment will be found, but it will be different from the unconstrained case, and possibly inconsistent with the other ligands. The point here is that the excluded volumes introduce an unnatural bias into the model building process that either causes you to lose some of the inactives altogether, or corrupts the information coming from them. So the net result is a loss of reliable information from low activity compounds, not a better QSAR model.

It is, however, perfectly reasonable to want to use excluded volumes when applying the model to screen out hits that occupy regions of space that are detrimental to activity (or at least predicted activity). One way to do this is to run Find Matches to Hypothesis *without* excluded volumes and let the hits align to the hypothesis and reference ligand in the most natural way. Then search those hits with the same hypothesis in the presence of excluded volumes, but with the "Score in place" option turned on. This will eliminate any of the naturally aligned hits that clash with the excluded volumes.

Keywords: Phase, pharmacophore, hypothesis, 3D QSAR, excluded volumes, Maestro

Back to Search Results

Was this information helpful?

What can we do to improve this information?

To ask a question or get help, please submit a support ticket or email us at
Knowledge Base Search

Type the words or phrases on which you would like to search, or click here to view a list of all
Knowledge Base articles