Quantitative structure-activity relationship QSAR methodology

Over the past 60 years a multitude of means have been used to capture SARs. We can broadly divide them into two groups those based on statistical or data mining methods (e.g., regression models) and those based on physical approaches (e.g., pharmacophore models). For a comprehensive review of quantitative structure-activity relationship (QSAR) methodologies the reader is referred... 

It could be argued that drug discovery strategies of the past were analogous to fishing with a line and hook. A systematic series of trial and error experiments such as quantitative structure-activity relationship (QSAR) methodologies, for example, were commonplace. The process was iterative and labor intensive. A chemist or biologist had relatively few restrictions, particularly with time, to explore ideas and test hypotheses. The overall endeavor was more craft than process. 

In a very recent publication,a new quantitative structure-activity relationship (QSAR) methodology multi-pH QSAR was proposed. This QSAR approach may be utilized to differentiate the activity of neutral and ionized species by considering two species-specific terms, log (neutral) and log D (ionized). 

Quantitative Structure—Activity Relationships (QSAR). Quantitative Stmcture—Activity Relationships (QSAR) is the name given to a broad spectmm of modeling methods which attempt to relate the biological activities of molecules to specific stmctural features, and do so in a quantitative manner (see Enzyme INHIBITORS). The method has been extensively appHed. The concepts involved in QSAR studies and a brief overview of the methodology and appHcations are given here. 

Certain computational methodologies such as some approaches to quantitative structure-activity relationship (QSAR) studies use 3D ligand structures [37, 38]. These methods generally assume that a bioactive conformation has been estab-Hshed for a set of molecules and that these conformers can be ahgned in a maimer that reflects the relative orientation they would adopt in a binding site. It is thus... 

Although the above methodologies proved to be very successful in identifying active kinase inhibitors, they utilized "generic" kinase models and did not address selectivity issues. An interesting recent report has attempted to create quantitative structure-activity relationship (QSAR) models based on data sets of compounds tested against multiple kinases [33]. 

Lipophilicity appears in several Quantitative Structure-Activity Relationships (QSAR) studies [16], emphasizing its importance. Different in vitro assays have been reported to measure lipophilicity from the classical shake-flask technique that still remains the reference for lipophilicity measurements to more actual methodologies. The first procedure is time-consuming, sensitive to impurities and the measurable log Poct range restricted to -3 to 3 [17]. 

New applications, new methodologies, and new perspectives are offered in this second volume. We have arranged the contributions as follows. First are four chapters dealing with conformational analysis, molecular mechanics, and molecular dynamics. Following these are four chapters on quantum mechanically oriented topics and two chapters on quantitative structure-activity relationships (QSAR). Lastly, an essay focusing on pivotal papers and trends in the computational chemistry literature and an updated appendix on software for molecular modeling are presented. 

The simplest methodology is two-dimensional (2D) quantitative structure-activity relationships (QSAR), in which calculated descriptors of molecules are related to an end point of interest via a mathematical relationship to estimate a numerical or categorical value for that end point. The mathematical relationship is fitted to a training set of compounds for which data for the end point has been measured experimentally. New molecules can then be described with the descriptors used in the model and their end point values predicted. 2D QSAR methods can be used to predict the interaction of compoimds with protein targets or antitargets and are widely used for prediction of physicochemical and ADME properties, such as hpophilicity, solubility, hiunan intestinal absorption, and blood-brain barrier penetration [18]. An excellent review of the strategies and pitfalls of 2D QSAR has been published by Lewis and Wood [19]. 

Lajiness MS (1991) Evaluation of the performance of dissimilarity selection methodology. In Silipo C, Vittoria A (eds) QSAR rational approaches to the design of bioactive compounds. Proceedings of the VIII European symposium on quantitative structure-activity relationships. Sorrento, Italy, 9-13 Sept 1990. ESCOM, Leiden, pp 201-204... 

---