This work demonstrated that similar conformers align to a reference shape in a similar way.
In the first step, a starting conformer is randomly selected as a reference shape.
If there is sufficient similarity to a reference shape, the corresponding binary bit is set.
Dubbed "alignment recycling", this approach recognized that conformers with similar shape align to a reference shape in a similar way.
In our case, the reference template is a reference shape pre-selected during the initial diverse selection.
The second and third steps are repeated until all conformers are assigned to a reference shape cluster.
Some conformers may have different optimal alignments with a reference shape due to structural symmetry; however, the presence of multiple reference shapes greatly increases the chance of finding an alignment very close to the analytical maximum overlap solution.
In contrast to ROCS, AR can only occur when the query conformer structure is found to have a reference shape in common with a database conformer.
If a reference shape cannot be found in common, the conformers are considered to be too different to be a neighbor.
The suitable Transform-Tanimoto value, which determines if two structures have a reference shape in common and enables alignment via matrix multiplication, was determined empirically.
