Table 1 Different 3-D similarity search (or analysis) scenarios considered

A

Compound

Single conformer

Similarity scores for a compound “query” compared to those of the “database” compounds, both computed using a single conformer per compound.

B

Compound

Multiple conformer, All-conformer-pair approach

Similarity scores that one may expect when each “query” conformer is compared to a set of multiple diverse conformers of the “database” compounds, using the “all-conformer-pair” approach. That is, all unique conformer pairs contribute to the average and standard deviation of the resulting similarity scores.

C

Conformer

Multiple conformer, All-conformer-pair approach

Similarity scores that one may expect when a single “query” conformer is compared to a set of multiple diverse conformers of the “database” compounds, using the “all-conformer-pair” approach. That is, all unique conformer pairs contribute to the average and standard deviation of the resulting similarity scores.

D

Conformer

Multiple conformer, Best-conformer-pair approach

Similarity scores that one may expect when a single “query” conformer is compared to a set of multiple diverse conformers of the “database” compounds using the “best-conformer-pair” approach. That is, only the conformer pair with the greatest similarity per conformer-compound pair contributes to the average and standard deviation of the resulting similarity scores.

E

Compound

Multiple conformer, Best-conformer-pair approach

Similarity scores that one may expect when each “query” conformer is compared to a set of multiple diverse conformers of the “database” compounds using the “best-conformer-pair” approach. That is, only the conformer pair with the greatest similarity per compound-compound pair contributes to the average and standard deviation of the resulting similarity scores.