Model-building

1 Model building. Classical QSAR analysis is based on two premises first, that biological activity of compounds and differences in potency are a function of 

However, in the first instance, QSAR attempts to express compound potency as a linear function of various structural and property descriptors D with coefficients weighting their relative importance  

The descriptor set can then be reduced by eliminating candidates that show such bad characteristics. Optimization techniques such as genetic algorithms (see Section 9.7) are powerful means of automating this selection process. 

On the other hand, techniques like Principle Component Analysis (PCA) or Partial Least Squares Regression (PLS) (see Section 9.4.6) are used for transforming the descriptor set into smaller sets with higher information density. The disadvantage of such methods is that the transformed descriptors may not be directly related to single physical effects or structural features, and the derived models are thus less interpretable. 

The model building step deals with the development of mathematical models to relate the optimized set of descriptors with the target property. Two statistical measures indicate the quality of a model, the regression coefficient, r, or its square, r, and the standard deviation, a (see Chapter 9). 

Model building consists of three steps training, evaluation, and testing. In the ideal case the whole training data set is divided into three portions, the training, the evaluation set, and the test set. A wide variety of statistical or neural network 

Fujita et al. were the first to develop a calculation method that was based, analogously to the Hammett approach, on substituent constants r[6] (see Eq. (7)) 

In accord with both theory and experience, the total function can be separated into a function of temperature and a function of the concentration variables  

This very general equation is the starting point for developing an analysis of process kinetics (see Chap. 4). 

Model building is concerned with the establishment of equations of this type. Important considerations in doing this are 

Identification with the model parameters is accomplished in most cases with use of graphically estimated linearizations or of nonlinear regression analysis. After model identification, parameterization may be undertaken. In the beginning the control of parameterization should not be delegated to the computer operating blindly with prescribed criteria. Deviations can suggest a previously unknown influence that should be incorporated into the model. 

A sensitivity-analysis has to show finally how the model reacts to changing values of parameters. An efficient biokinetic experimental design has been demonstrated by Johnson and Berthouex (1975a), and multiresponse data are used for parameter estimation by the same authors (1975b). 

We adopted the method by Car and Parrinello [22] to ensure a self-consistent evolution of the electronic structure during molecular dynamics motion. The electronic structure was described in the framework of density functional theory (DFT) with the generalized gradient approximation (GGA) due to Becke (B) for the exchange energy and Lee, Yang and Parr (LYP) for the correlation energy 

---

It is very important, from one hand, to accept a hypothesis about the material fracture properties before physical model building because general view of TF is going to change depending on mechanical model (brittle, elasto-plastic, visco-elasto-plastic, ete.) of the material. From the other hand, it is necessary to keep in mind that the material response to loads or actions is different depending on the accepted mechanical model because rheological properties of the material determine type of response in time. The most remarkable difference can be observed between brittle materials and materials with explicit plastic properties. 

Friedman S H, DeCamp D L, Si]besma R, Srdanov G and WudI F 1993 Inhibition of HIV-1 protease by fullerene derivatives model building studies and experimental verifioation J. Am. Chem. See. 115 6506-9... 

Thi.s method of cstabli.shing a rclation.ship between a molcculai structure and it.s properties is inductive. It depends on a set of compounds with know n properties or activities whicli is used for model building. 

Iteration of the steps, descriptor selection, model building, and model validation in combination with an optimi ation algorithm allows one to select a descriptor subset having maximum predictivity. 

The method of building predictive models in QSPR/QSAR can also be applied to the modeling of materials without a unique, clearly defined structure. Instead of the connection table, physicochemical data as well as spectra reflecting the compound s structure can be used as molecular descriptors for model building,... 

A structure descriptor is a mathematical representation of a molecule resulting from a procedure transforming the structural information encoded within a symbolic representation of a molecule. This mathematical representation has to be invariant to the molecule s size and number of atoms, to allow model building with statistical methods and artificial neural networks. 

The abbreviation QSAR stands for quantitative structure-activity relationships. QSPR means quantitative structure-property relationships. As the properties of an organic compound usually cannot be predicted directly from its molecular structure, an indirect approach Is used to overcome this problem. In the first step numerical descriptors encoding information about the molecular structure are calculated for a set of compounds. Secondly, statistical methods and artificial neural network models are used to predict the property or activity of interest, based on these descriptors or a suitable subset. A typical QSAR/QSPR study comprises the following steps structure entry or start from an existing structure database), descriptor calculation, descriptor selection, model building, model validation. 

To become familiar with the application of the basic principles of the model building process by means of calculating log P and log 5 values... 

The general procedure in a QSPR approach consists of three steps structure representation descriptor analysis and model building (see also Chapter X, Section 1.2 of the Handbook). 

Figure 10.1-1. Flow chart for the general model building process in QSPR studies.

The AMBER (Assisted Model Building and Energy Refin emeni) is based on a force field developed for protein and nucleic acid computations by members of the Peter Kollman research group at the... 

Jones T A and S Thirup 1986. Using Known Substructures in Protein Model Building and Crystallography. EMBO journal 5 819-822. 

Assisted model building with energy refinement (AMBER) is the name of both a force field and a molecular mechanics program. It was parameterized specifically for proteins and nucleic acids. AMBER uses only five bonding and nonbonding terms along with a sophisticated electrostatic treatment. No cross terms are included. Results are very good for proteins and nucleic acids, but can be somewhat erratic for other systems. 

AMBER (assisted model building with energy refinement) a molecular mechanics force field... 

As you go through this text you will see two dif ferent modeling icons The SpartanBuild icon alerts you to a model building opportunity the SpartanView icon indicates that the Learning By Modeling CD includes a related model or animation... 

The ball and wire display is used for model building Although it is convenient for this purpose other model displays show three dimensional molecular structure more clearly and may be preferred The space filling display is unique m that it portrays a molecule as a set of atom centered spheres The individual sphere radii are taken from experi mental data and roughly correspond to the size of atomic electron clouds Thus the space filling display attempts to show how much space a molecule takes up... 

In collaboration with Wavefunction we have created a cross function CD ROM that contains an electronic model building kit and a rich collection of molecular models that reveal the interplay between electronic struc ture and reactivity m organic chemistry... 

P. A. KoUman and co-workers, AMBER Assisted Model Building and Energy Refinement, Urdversity of CaUforrda, San Erancisco, 1980—present ... 

Cropley, J.B., Systematic Errors in Recycle Reactor Kinetic Studies, Chemical Engineeiing Piogiess, February 1987, 46-51. (Model building, experimental design)... 

The vertices are connected with hues indicating information flow. Measurements from the plant flow to plant data, where raw measurements are converted to typical engineering units. The plant data information flows via reconciliation, rec tification, and interpretation to the plant model. The results of the model (i.e., troubleshooting, model building, or parameter estimation) are then used to improve plant operation through remedial action, control, and design. 

Identifying the minimum number of specific measurements containing the most information such that the model parameters are uniquely estimated requires that the model and parameter estimates be known in advance. Repeated unit tests and model building exercises will ultimately lead to the appropriate measurements. However, for the first unit test in absence of a model, the identification of the minimum number of measurements is not possible. 

However, given that reconciliation will not always adjust measurements, even when they contain large random and gross error, the adjustments will not necessarily indicate that gross error is present. Further, the constraints may also be incorrect due to simphfications, leaks, and so on. Therefore, for specific model development, scrutiny of the individual measurement adjustments coupled with reconciliation and model building should be used to isolate gross errors. 

---