The predictive power of 14 calculation procedures for molecular lipophilicity is checked by comparing with reliable experimental logP values from the literature. The database of 138 test compounds comprises 90 simple organic structures and 48 chemically heterogeneous drug molecules (P-blockers, class I antiarrhythmics and neuroleptics).The present investigations lead us to conclude that the predictive power of the calculation procedures is significantly better for simple organic molecules than for chemically heterogeneous drug structures. The calculation procedures should be arranged in three groups with significantly differing predictive power: fragmental > atom-based > conformation-dependent approaches.

No abstract

No abstract

In the ®rst part of this paper we brie¯y describe experimental (octanolywater partitioning) as well as computational approaches to quantifying molecular lipophilicity. The central section focuses on the hydrophobic fragmental constant approach (Sf -system) as developed by Rekker and his group, starting in the early seventies. The original approach has been extended and revised a number of times; the most recent updating is presented here. It is followed by a detailed description of how to apply the correction factor C M . The practical procedure of Sf-calculations is described for some examples and the validity of these calculations is veri®ed by comparison with other calculation methods and experimental data.

Experimental and calculated data of drug lipophilicity, i.e. octanol/water partition coefficients (log Pobs) and hydrophobic fragmental constants (Σf, C log P) have been compared for the following four groups of drugs: antiarrhythmics (AA), β‐blockers (BB), phenothiazines (PT) and benzamides (BA). Measured log P values indicate significantly varying lipophilicity ranges in the case of these four groups (AA>BB>BA>PT). Comparison of the two calculative approaches with observed log P values yields almost identical qualities. Strongest discrepancies between calculative and experimental data are discussed on the basis of putative experimental pitfalls or in relation to structurally derived miscalculations.