Phase Hypothesis Identification

The Phase hypothesis identification node runs sequentially pharmacophore site creation, common pharmacophore perception and hypothesis scoring.

A set of pharmacophore feature definitions is applied to each ligand conformation to identify the positions of all pharmacophoric elements in that ligand.

During common pharmacophores identification all n-point pharmacophores from the input ligands are enumerated and filtered into a set of high-dimensional boxes. Pharmacophores that fall into the same box are similar enough to be considered equivalent. Boxes that receive at least one pharmacophore from a sufficient number of actives are said to "survive" the partitioning process. Each surviving box may be processed in the subsequent Score Actives step to identify a pharmacophore hypothesis from that box. Hypotheses are scored with respect to input ligands by assigning numerical rankings to the pharmacophores within each surviving box. The highest scoring pharmacophore in a given box is designated as a hypothesis, and the ligand giving rise to that pharmacophore is known as its reference ligand. The scoring function considers the quality of the alignments, of the oriented feature overlay and of the conformation volume overlap.

Backend implementation