Lateral validation

Consistent with other analytical methods, immunoassays must be validated to ensure that assay results are accurate. Initial validation involves an evaluation of the sensitivity and specificity of the immunoassay, while later validation includes comparison with a reference method. Because a goal of immunoassays is to minimize sample preparation, validation also includes testing the effects of sample matrices and(or) sample cleanup methods on results. The final steps in validation involve testing a limited number of samples containing incurred residues to determine if the method provides reliable data. 

In the 1990s, ECVAM held a forum to vet and evaluate new alternative assays, and developed a list of compounds for testing (24). The key driver for this activity was the fact that DART studies require large numbers of animals. The primary focus of this activity was embryo-fetal toxicity. The list generated from this forum was tested in three assays (later validated by ECVAM) (1) the micromass assay, (2) the rat WEC assay, and (3) the embryonic stem cell test (25). Compounds on the Brown list were classified as either strong, weak, or non-teratogens. The three assays successfully predicted the compound classification about 80% of the time. However, the embryonic stem cell test later performed poorly on a different group of chemicals with known in vivo activities (26). 

The appearance of the o-" parameter in a large number of reactions and interactions involving X-phenols indicates that the phenoxy radical can be a potent, reactive intermediate in myriad reactions. The availability of a fast, easily retrievable computerized database to corroborate this phenomenon was useful. This approach of lateral validation was crucial in establishing a QSAR model that was not only statistically significant but also mechanistically interpretable. 

Laplacian spectrum - Laplacian matrix lateral validation -> validation techniques... 

Kim, K.H. (1995b). Comparison of Classical QSAR and Comparative Molecular Field Analysis Toward Lateral Validations. In Classical and Three-Dimensional QSAR in Agrochemistry (Hansch, C. and Fujita, T, eds.), American Chemical Society, Washington (DC), pp. 302-317. 

The methods may be done in blocks, with the later validation steps, such as robustness being the icing on the cake of the method. In large companies, decisions will be made at different stages to continue work on the method and validation, or to cease development. Some aspects of the validation process are discussed below. 

Kim KH. Comparison of classical QSAR and comparative molecular field analysis Toward lateral validations. In Hansch C, Fujita T, eds. Classical and Three-Dimensional QSAR In Agro-Chemistry, ACS Symposium series. Vol. 606. Washington, DC American Chemical Society, 1995 302-317. 

Except for a very few studies most of the work so far has been done using SAR data of wild-type protease. Further QSAR and molecular modeling studies on prodrug derivatives, wild-type vs. mutant SAR data and lateral validation of these models via comparative analysis can provide useful insight. Such studies can highlight the differences and similarity, if any, in their mechanism of interaction with wild-type and mutant protease receptor. 

The concept of lateral validation was first formulated by Hansch for classical QSARs. In this approach, the choice of parameters, their sign, and the size of their coefficients are compared with those from other QSARs. A comparison is illustrated in Table 5 for the Hammett equation ... 

Two of the QSARs in Table 5 were derived on five compounds apiece but the consistency of the QSARs validate the results. Thus, if the same parameter is present, with a similar contribution, in the QSAR analyses being compared, more confidence can be attributed to all the models. By doing comparative analyses on multiple systems, rather than single models, lateral validation allows deeper understanding of the studied problem in its complexity. Several cytochrome P450 classical QSARs have already been analyzed with this method. 

Lateral validation of CoMFA models with classic QSAR models was performed by comparing the informative content of PLS components (from CoMFA) with classic QSAR descriptors (e.g., Hammett s ct constant) for the dissociation constants of benzoic and phenylacetic acids. This approach, useful in comparing CoMFA to classical QSARs, could be routinely used in CoMFA studies. Kim s work illustrates the principle that classical QSAR should always be performed to complement CoMFA studies (see also final remarks). 

Table 5 Example of Lateral Validation for Classical QSARs "... 

Lateral validation of 3D-QSARs is a meta-analytical technique allowing the user to design molecules that would have specificity for one, but not other receptors, revealing differences and similarities between the targeted receptors or between the compared series as systems, not as individual compounds. Because of the underlying assumption that the compared QSAR models are correct, this technique cannot guarantee robustness or predictive and explanatory power. 

Figure 7 Qualitative lateral validation of steroid-based QSARs derived for aromatase (Ref. 96), DB3, (Ref. 110), and the androgen (Ref. 92) and estrogen (Ref. 130) receptors, shown for the steric fields only.

Aiming at later validation, hats of all four materials were exposed in an IP/DP box in Phoenix/Arizona see Fig. 14.1. This exposure site was chosen, as it features high global solar radiation combined with relative dry climate. Besides avoiding... 

Hansch recommended a lateral validation of QSAR results, i.e., the comparison of models of closely related series of compounds in one biological test system (cf. equations 16 and 17) or trie comparison of the QSAR models derived for one series of compounds in several related biological test models (e.g., serine and cysteine proteases). If all models are of comparable quality, and if they show similar regression coefficients of the physicochemical terms, the results can be accepted. However, in most cases the required effort will be too large to do this routinely. In addition, even closely related enzymes or receptors may have significantly different binding sites. 

---