As the use of Quantitative Structure Activity Relationship (QSAR) models for chemical management increases, the reliability of the predictions from such models is a matter of growing concern. The OECD QSAR Validation Principles recommend that a model should be used within its applicability domain (AD). The Setubal Workshop report provided conceptual guidance on defining a (Q)SAR AD, but it is difficult to use directly. The practical application of the AD concept requires an operational definition that permits the design of an automatic (computerised), quantitative procedure to determine a model's AD. An attempt is made to address this need, and methods and criteria for estimating AD through training set interpolation in descriptor space are reviewed. It is proposed that response space should be included in the training set representation. Thus, training set chemicals are points in n-dimensional descriptor space and m-dimensional model response space. Four major approaches for estimating interpolation regions in a multivariate space are reviewed and compared: range, distance, geometrical, and probability density distribution.
JaworskaJ., ComberM., Van LeeuwenC., & AuerC. (2003). Summary of the workshop on regulatory acceptance of QSARs. Environmental Health Perspectives111, 1358–1360.
3.
TunkelJ., MayoK., AustinC., HickersonA., & HowardP. (2005). Practical considerations on the use of predictive models for regulatory purposes. Environmental Science and Technology39, 2188–2199.
4.
Sheridan, R., Feuston, R.P., Maiorov, V.N. & Kearsley, S. (2004). Similarity to molecules in the training set is a good discriminator for prediction accuracy in QSAR. Journal of Chemical Information and Computer Science44, 1912–1928.
5.
Nikolova, N. & Jaworska, J. (2002). Review of chemical similarity measures. QSAR & Combinatorial Science22, 1006–1026.
6.
Bender, A. & Glen, R.C. (2004). Molecular similarity: a key technique in molecular informatics. Organic and Biomolecular Chemistry2, 3204–3218.
7.
Kubinyi, H. (1998). Similarity and dissimilarity: A medicinal chemist's view. Perspectives in Drug Discovery and Design9-11, 225–252.
8.
Martin, Y.C., Kofron, J.L. & Traphagen, L.M. (2002). Do structurally similar molecules have similar biological activity?Journal of Medicinal Chemistry45, 4350–4358.
9.
Debnath, A.K., Debnath, G., Shusterman, A.J. & Hansch, C. (1992). A QSAR investigation of the role of hydrophobicity in regulating mutagenicity in the Ames test: 1. Mutagenicity of aromatic and heteroaromatic amines in Salmonella typhimurium TA98 and TA100. Environmental Molecular Mutagenesis19, 37–52.
10.
Glende, C., Schmitt, H., Erdinger, L., Engelhardt, G. & Boche, G. (2001). Transformation of mutagenic aromatic amines into non-mutagenic species by alkyl substituents. Part I. Alkylation ortho to the amino function. Mutation Research498, 19–37.
11.
TOPKAT OPS (2000). US patent no. 6 036 349 issued March 14, 2000.
12.
TongW., QianX., HuixiaoH., LemingS., HongF., & PerkinsR. (2004). Assessment of prediction confidence and domain extrapolation of two Structure– Activity Relationship models for predicting estrogen receptor binding activity. Environmental Health Perspectives112, 1249–1254.
13.
SeberG.A.F. (1984). Multivariate Observations.712 pp., New York, NY, USA: John Wiley & Sons.
14.
Nikolova-JeliazkovaN., & JaworskaJ. (2005). An approach to determining AD for QSAR group contribution models: an analysis of SRC KOWWIN. ATLA33, 461–470.
15.
PreparataF.P., & ShamosM.I. (1991). Convex hulls: Basic algorithms. In Computational Geometry: An Introduction (ed. PreparataF.P., & ShamosM.I.), pp. 95–148, New York, NY, USA: Springer-Verlag.
16.
FukunagaK. (1990). Introduction to Statistical Pattern Recognition. 2nd edn, 592 pp. Computer Science and Scientific Computing Series. New York, NY, USA: Academic Press.
17.
TropshaA., GramaticaP., & GombarV. (2003). The importance of being earnest: validation is the absolute essential for successful application and interpretation of QSPR Models. QSAR & Combinatorial Sciences22, 69–77.
18.
ErikssonL., JaworskaJ., WorthA., CroninM.T.D., McDowellR.M., & GramaticaP. (2003). Methods for reliability and uncertainty assessment and for applicability evaluations of classification-and regression-based QSARs. Environmental Health Perspectives111, 1351–1375.
19.
NeterJ., KutnerM.H., WassermanW., & NachtsheimC. (1996). Applied Linear Statistical Models.1408 pp. New York, NY, USA: McGraw-Hill.
20.
MyersR.H. (2000). Classical and Modern Regression with Applications. 2nd edn, 488 pp. Pacific Grove, CA, USA: Duxbury Press.
21.
SilvermanB.W. (1986). Density Estimation for Statistics and Data Analysis.176 pp. Boca Raton, FL, USA: CRC Press.
22.
ECETOC. (2003). Evaluation of commercially available software for human health, environmental and physico-chemical endpoints. ECETOC QSAR TF report nr 89, 164 pp. Brussels, Belgium: ECETOC.
23.
NetzevaT.N., WorthA.P., AldenbergT., BenigniA., CroninM.T.D., GramaticaP., JaworskaJ.S., KahnS., KlopmanG., MarchantC., MyattG., Nikolova-JeliazkovaN., PatlewiczG.Y., PerkinsR., RobertsD., SchultzT.W., StantonD., van de SandtJ.J., TongW., VeithG., & YangC. (2005). Current status of methods for defining the applicability domain of (Quantitative) Structure-Activity Relationships. The report and recommendations of ECVAM Workshop 52. ATLA33, 1–19.
FriedmanJ.H. (1987). Exploratory projection pursuit. Journal of the American Statistic Association82, 249–266.
28.
ChenM.-H., & ShaoQ-M. (1999). Monte Carlo Estimation of Bayesian Credible and HPD Intervals. Journal of Computational and Graphical Statistics8, 69–92.
PressW., TeukolskyS., VetterlingW., & FlanneryB. (1992). Numerical Recipes in C: The Art of Scientific Computing, 2nd edn, 1020 pp. Cambridge, UK: Cambridge University Press.
