SUMMARY: number of rotatable bonds in drug-like molecules: references ?

 Thanks very much to all who responded.
 The paper by Oprea says that 70% of drug-like molecules have between 2 and 8
 rotatable bonds.
 My understanding of the rationale behind this number is that rotation around
 8-9 rotatable bonds would cost ~ 9 * 1.4 kj/mol/rotatable bond = 12.6
 kJ/mol/rotatable bonds, or ~ 3kcal/mol/rotatable bond. This would be
 commpensated by a hydrogen bond when the ligand is bound to the receptor.
 Rotattion around a higher number of rotatable bonds would be more difficult
 to compensate in term of non-bonded ligand/receptor interactions (?)
 Commercial softwares include this descriptor as 'drug like'. MOE has a
 default value of 5 rotatable bonds. MSI' Cerius2 lists the # of rotatable
 bonds in the same structural subset  of the descriptor table than the
 rule-of-5 descriptors, but does not propose  a default value. I don't know
 about Tripos or OxMol or ... (additional informations welcome).
 Here is the summary of the answers I get: (!!!! ADDITIONAL ANSWERS ARE
 a good reference is: T. Oprea J. Comp.-Aided Mol. Design 14:251-264 (2000)
 Property distribution of drug-related chemical matter. According to Opreas
 analysis, 70% of drug-like compounds have between 2 and 8 rotatable bonds.
 Hope this helps,
 Ingo Muegge
 You might take a look at a recent paper by T. Oprea (Property Distribution
 in Drug-related database, J. Comput. Aided Mol Des., 2000, 14(3):251-264)
 where he describes about number of rotatable bonds (RTB) and rigid bonds
 (RGB), etc.
     Hope this helps.
 Asim Debnath
 See Bohm's paper, "The Development of a simple empirical scoring
 function to estimate the binding constant for a protein-ligand complex
 of known three-dimensional structure" Journal of Computer-Aided
 Molecular Design, Vol 8, p 243-256 (1994) in which he finds a cost of
 1.4 kJ/mol/rotatable bond.
 In a similar vein, see the paper by M. D. Eldridge et. al, "Empirical
 Scoring functions: I. The development of a fast empirical scoring
 function to estimate the binding affinity of ligands in receptor
 complexes" Journal of Computer-Aided Molecular Design, Vol 11 p. 425-445
 (1997) where a slightly larger coefficient for a slightly different
 functional form than Bohm's is found.
 Kristy Mardis,
 Jerome Baudry, Ph.D.
 Research Scientist, Computational Chemistry
 TransTech Pharma, Inc.
 4170 Mendenhall Oaks Pwky, Suite 110
 High Point, NC, 27265
 jbaudry ^at^
 tel: (336) 841-0300 #121 or #120
 fax: (336) 841-0310