Domain of applicability

In contrast to these scientific issues, we can look back to a prediction made by Snyder in 1991, at the start of the age of modelling. He said that the greatest impact of molecular modelling has come, and will continue to come, from computational chemists who are deeply embedded in the drug discovery process as full partners to the project teams. Despite the advances that have been made in theory, and the increase in computer power, this human factor is still the key to successful modelling. 

Naylor, I. Motoc and G. R. Marshall, J. Comput. -Aided 

Cheney, J. S. Mason, S. Maignan, J. P. Guilloteau, K. Brown, D. Colussi, 

Guba and M. Kansy, Drug Discovery Today, 2006, 11, 326. 

Dresslar, J. Nerothin and H. A. Carlson, Nucleic Acids Res., 2008, 36, D674. 

---

In this section we present several numerical teclmiques that are conmronly used to solve the Sclirodinger equation for scattering processes. Because the potential energy fiinctions used in many chemical physics problems are complicated (but known to reasonable precision), new numerical methods have played an important role in extending the domain of application of scattering theory. Indeed, although much of the fomial development of the previous sections was known 30 years ago, the numerical methods (and computers) needed to put this fomialism to work have only been developed since then. 

Every phase of internal coordinate modeling admits many methodological variations, and I do not attempt to review them all. I outline only the standard problems encountered in any particular domain of application and the common practical solutions. 

The domain of applicability of some of the most crucial postulates is not clearly established. We examine these aspects in detail. 

The work described in this paper is an illustration of the potential to be derived from the availability of supercomputers for research in chemistry. The domain of application is the area of new materials which are expected to play a critical role in the future development of molecular electronic and optical devices for information storage and communication. Theoretical simulations of the type presented here lead to detailed understanding of the electronic structure and properties of these systems, information which at times is hard to extract from experimental data or from more approximate theoretical methods. It is clear that the methods of quantum chemistry have reached a point where they constitute tools of semi-quantitative accuracy and have predictive value. Further developments for quantitative accuracy are needed. They involve the application of methods describing electron correlation effects to large molecular systems. The need for supercomputer power to achieve this goal is even more acute. 

U-series nuclides define their domain of application the longest-lived ( °Th, " " Pa, Ra), as discussed above, are appropriate to study partial melting in the mantle or deep crustal reservoirs, the shortest-lived give information on processes occurring in shallow magma chambers. 

Thus, the domain of applicability of the first type of polycrystalline adsorbents with intercrystalline threshold contacts is provided by the common gaseous analysis of air media. Accurate vacuum studies and... 

Thus, we have considered in detail various theoretical models of effect of adsorption of molecular, atom and radical particles on electric conductivity of semiconductor adsorbents of various crystalline types. Special attention has been paid to sintered and partially reduced oxide adsorbents characterized by the bridge type of intercrystalline contacts with the dominant content of bridges of open type because of wide domain of application of this very type of adsorbents as sensitive elements used in our physical and chemical studies. 

This hybrid approach can significantly extend the domain of applicability of the AIMS method. The use of interpolation significantly reduces the computational effort associated with the dynamics over most of the timescale of interest, while regions where the PESs are difficult to interpolate are treated by direct solution of the electronic Schrodinger equation during the dynamics. The applicability and accuracy of the method was tested using a triatomic model collisional quenching of Li(p) by H2 [125], which is discussed in Section III.A below. 

It needs to be emphasized that no matter how robust, significant, and validated a QSAR may be, it cannot be expected to reliably predict the modeled property for the entire universe of chemicals. Therefore, before a QSAR model is put into use for screening chemicals, its domain of application must be defined and predictions for only those chemicals that fall in this domain should be considered reliable. Some approaches that aid in defining the applicability domain are described below. 

In closing, I want to stress again the essential importance of understanding in a fundamental way the nonlocal structure of exchange-correlation contributions to energy functionals. I believe that a full appreciation of the variety of ways that this nonlocal structure manifests itself, according to the different physical circumstances, will be vital for the construction of improved representations of F[n] as the domain of applications of density functional theory continues to be extended into exciting new areas. 

In 2008, Weaver [64] utilized PPB as an example to demonstrate the concept of "domain of applicability" in QSAR researches. The PLS model was constructed using 17 ID and 2D molecular descriptors. The performance of the model was reasonable for such a large data set for PPB modeling (n — 685, q2 — 0.56, RMSE = 0.55 AUE = 0.42, ntest = 210, q2 = 0.58, RMSEtest = 0.54, AUEtest = 0.41). How domain selection protocol affects the prediction performance will be discussed in Section 3. 

Once a QSAR model is constructed, it must be validated using the external test set. The data points in the test set should not appear in the training set. There are two approaches to improve the prediction accuracy for a given QSAR model. The first approach utilized the concept of "the domain of applicability," which is used to estimate the uncertainty in prediction of a particular molecule based on how similar it is to the compound used to build the model. To make a more accurate prediction for a given molecule in the test set, the structurally similar compounds in the training set are used to construct model and that model is used to make the prediction. In some cases, the domain similarity is measured using molecular descriptor similarity, rather than the structural similarity. The... 

It is important to note that if generic problem formulations are to be used, it is especially important that they are developed very carefully in the first place, that their domains of applicability are carefully defined, and that users should doublecheck on every occasion that they are fully appropriate to the case in hand or adjust them as necessary. 

Weaver, S. and Gleeson, M.P. (2008) The importance of the domain of applicability in QSAR modeling. Journal of Molecular Graphics e[ Modelling 26, 1315. 

Figure 5.2 The domain of application of in vitro techniques used to measure partition coefficients in 1-octanol/water system and in other diphasic systems. In gray medium- or low-throughput methods not described in this chapter.

Schroeter, T, Schwaighofer, A., Mika, S., Ter Laak, A., Suelzle, D., Ganzer, U., Heinrich, N. and Muller, K.R. (2007) Machine learning models for lipophilicity and their domain of applicability. Molecular Pharmaceutics, 4, 524—538. 

Finally, we present the results of the case studies for Eley-Rideal and LH reaction mechanisms illustrating the practical aspects (i.e. convergence, relation to classic approximations) of application of this new form of reaction rate equation. One of surprising observations here is the fact that hypergeometric series provides the good fit to the exact solution not only in the vicinity of thermodynamic equilibrium but also far from equilibrium. Unlike classical approximations, the approximation with truncated series has non-local features. For instance, our examples show that approximation with the truncated hypergeometric series may supersede the conventional rate-limiting step equations. For thermodynamic branch, we may think of the domain of applicability of reaction rate series as the domain, in which the reaction rate is relatively small. 

One of the most important problems in QSAR analysis is establishing the domain of applicability for each model. In the absence of the applicability domain restriction, each model can formally predict the activity of any compound, even with a completely different structure from those included in the training set. Thus, the absence of the model applicability domain as a mandatory component of any QSAR model would lead to the unjustified extrapolation of the model in the chemistry space and, as a result, a high likelihood of inaccurate predictions. In our research we have always paid particular attention to this issue (12, 20-27). A good overview of commonly used applicability domain definitions can be found in reference (28). 

Oherg, T. (2004) A QSAR for baseline toxicity validation, domain of application, and prediction. Chem. Res. Toxicol.,... 

The strong points of this technique are its absence of interaction with the stationary phase, rapid dilution and the potential to recover all of the analytes. Because the technique permits the separation of nominal masses ranging from 200 to 107 Da, its main applications are in the analyses of synthetic and natural polymers. The choice of stationary phase for a given separation is made by examination of the calibration curve of various columns. The column of choice is that which provides a linear range over the masses of the compounds found in the sample. The calibration has to be conducted with the same type of polymers because macromolecules can have various forms ranging from pellet-like to thread-like. The data presented in Table 7.1 show the domains of application of the three gels shown in Fig. 7.3, depending on the standards that are used. 

In conclusion of this chapter let us discuss briefly the use of model (effective) potentials. They represent one more variety of simplified theoretical or semi-empirical calculations. Let us notice at once that these methods have a rather narrow domain of applicability. They are most frequently... 

Efficient methods of selection of admixed configurations in this approach are described in [12]. The computer code for generating the reduced configuration state list is presented in [13]. A special diagonalisation method is developed to reduce the orders of matrices to be diagonalised [14]. A more detailed overview of the methods considered may be found in [15]. All these improvements considerably speed up the calculations and widen the domains of applicability of the methods developed. 

---