Name: Create and Score Combinatorial Library from Known Estrogen Antagonists

Author: Ton van Daelen

Version: 2.0

Created: 9/2002

Modified: 8/10/2007

Purpose:
This protocol first builds a statistical model for a set of 116 known estrogen-antagonists. It then finds the maximum common subgraph present in at least 40% of the EA's. The "Generate R-groups" component uses this core to strip off all R-groups from the EA's containing that core. The core and R-groups are then enumerated to form a combinatorial library of ~20,000 compounds. The original EA's are mixed back in and all compounds are scored using the model. An html page shows known EA's mixed with a set of novel compounds that are predicted to have a high probability of exhibiting EA activity.

Requirements: Pipeline Pilot 6.0 or later
Chemistry
Data Modelling

O/S: Windows and Linux

Limitations: none known

Keyword: Bayesian learning modeling Rgroup MCSS core attachements

Contents: Create and Score Combinatorial Library from Known Estrogen Antagonists.xml

Installation:

1. Unzip the archive.
2. Import the protocol into the Pipeline Pilot client by dragging and dropping it in the white workspace area.
3. Run the protocol.