Tuplets

Overview

Tuplets™ enables researchers to retrieve compounds from molecular structure databases that are pharmacophorically, and therefore biologically, similar to known actives without the need to derive a classical 3D pharmacophore model. Structure databases are searched with pharmacophore fingerprint hypotheses (Tuplet hypotheses). Tuplet hypotheses can be derived from known active compounds, combined sets of active and inactive compounds, or UNITY^® queries. Compounds retrieved in this way can be of a different chemical class than those employed to generate the hypothesis, demonstrating Tuplets' lead hopping capabilities. Also, Tuplets can be used to prioritize virtual combinatorial libraries for further research.

Tuplets Brochure (765k)

A histogram (top left) is used to show the results of comparing a hypothesis (selected in the Tuplets - Hypothesis Manager - bottom center) to a reference set of 1,000 randomly selected drug-like molecules (shown in red) and the training set used to create the hypothesis (shown in green). The Tuplets - Analysis interface (top right) displays the number of compounds from both the training and reference sets that have a similarity greater than or equal to the specified cut-off value, allowing the user to interactively explore and apply multiple cut-off values for one or more similarity measurements.

Key Benefits

• Searching a molecular structure database using Tuplets is orders of magnitude faster than traditional 3D searching
• Tuplet hypotheses can be generated from HTS hits, well characterized active and inactive compounds, and/or UNITY queries
• Tuplets allows categorization of compounds by pharmacophoric properties
• Compound sets retrieved from molecular structure databases are enriched in actives
• Tuplets allows characterization of queries by comparison with reference data sets
• Tuplets can be used without the need to derive a classical 3D pharmacophore model