pharmacophore

1
pharmacophore

• Megon Walker
• Bioinformatics Program
• Boston University

2
prioritization of chemical libraries

• high-throughput screening

• Proceedings of the Fifteenth International
  Conference on Machine Learning, 19981-9.
• J. Chem Inf. Comput. Sci. 2003, 43 667-673.

3
compound datasets

• NIH/NCI developmental therapeutics program (DTP)
• human tumor cell line screen
• cancer screen of 43,000 compounds
• compounds tested at 5 different concentrations
  for the ability to inhibit 60 different human
  tumor cell lines
• dose response data
• 3 concentration parameters
• GI50
• TGI
• LC50
• MDL's SDFile format (2D structure)
• http//dtp.nci.nih.gov/docs/cancer/cancer_data.htm
  l

• DuPont KDD thrombin dataset
• 632 DuPont thrombin-targeting compounds
• 149 actives
• 483 inactives
• a binary feature vector for each compound with 1
  of 2 class labels
• shaped-based features
• pharmacophore features
• 139,351 features
• J Mol Graph Model. 2002 Jun20(6)469-77.
• J Chem Inf Comput Sci. 2002 Sep-Oct42(5)1230-40.
  
• KDD Cup 2001. http//www.cs.wisc.edu/dpage/kddcup
  2001/

4
pharmacophore derivation

• J Med Chem. 2004 Dec 247(25)6144-59.

5
pharmacophore derivation

• Curr Med Chem. 2004 Jan11(1)71-90.

6
pharmacophore derivation

• J Med Chem. 1999 Aug 2642(17)3251-64.
• Curr Med Chem. 2004 Jan11(1)71-90.

7
pharmacophore and screening

• Biogen
• bit string representation of 3D binding
  interactions used to filter compounds during
  virtual screening (docking)
• J Med Chem. 2004, 47 337-344.
• Deltagen
• machine learning and information theory to
  identify pharmacophore ensembles for compound
  classification
• J Chem Inf Comput Sci. 2003, 43 47-54
  2163-2169.
• Novartis
• naïve Bayes classifier for ranking enrichment
  of HTD/HTS results
• J Med Chem. 2004, 47 2743-2749, 4356-4359,
  6144-6159.
• J Chem Inf Comput Sci. 2004, 44993-9.