Prediction of Affinity

Head et al. developed a PLS-based model VALIDATE [47] to scale the relative contributions of entropy and enthalpy to binding affinity for a variety of complexes whose crystal structures had been determined. Molecular mechanics were used to calculate several parameters most correlated with enthalpy of binding, while changes in surface area, number of rotatable bonds fixed upon binding and other parameters more related to the entropy of binding were also included in the model. Of interest was that the principal components of the model were dominated by two terms (AH and AS, 

---

The different spatial position of the HY1-RA features (relative to the HY1-HY2 features of the first model for ald-AR antagonists) allowed a better prediction of affinity values for compounds bearing a para substituent at the phenyl ring attached to the piperazine nucleus. 

In summary, a large variety of methods have been deployed to predict ER binding of possible endocrine-disrupting chemicals. Consistently measured data for a large inventory of structurally diverse compounds are available [102,103] that will make extensive validation of the methods in future applications possible and also mandatory before helping regulatory bodies in the risk-assessment process. This validation should also include the prediction of affinities for compound classes that are not part of the training process. 

In this paper, the BPDA-PFMB/PEI molecular composites were oriented by means of zone annealing/drawing slightly above the glass transition temperatures of the respective molecular composites (280 - 400 °C). The dependence of draw ratio on tensile modulus, crystal orientation, and birefringence was determined as a function of composition. The relationship between structure (crystal orientation) and tensile property (modulus) of drawn films has been examined by comparing crystal chain orientations with the prediction of affine deformation (12-16). 

The critical factor for any method involving an approximation or an extrapolation is its range of application. Liu et al. [15] demonstrated that the approach performed well for mutations involving the creation or deletion of single atoms. The method has also been successfully applied to the prediction of the relative binding affinities of benzene, toluene and o-, p-, and m-xylene to a mutant of T4-lysozyme [16]. In both cases, however, the perturbation to the system was small. To investigate range over which the extrapolation may... 

RH Smith Jr, WL Jorgensen, J Tirado-Rives, ML Lamb, PAJ Janssen, CJ Michejda, MBK Smith. Prediction of binding affinities for TIBO inhibitors of HIV-1 reverse transcriptase using Monte Carlo simulations m a linear response method. J Med Chem 41 5272-5286, 1998. 

Predicted proton affinities of azoles (and oxazoles) calculated with simple ab initio methods (STO-3G) are reported to differ little from 6-3IG values (89KGS508). 

Equation 6.19 predicts an increasing IC50 with either increases in L or 1. In systems with low-efficacy inverse agonists or in systems with low levels of constitutive activity, the observed location parameter is still a close estimate of the KB (equilibrium dissociation constant of the ligand-receptor complex, a molecular quantity that transcends test system type). In general, the observed potency of inverse agonists only defines the lower limit of affinity. 

Equation (32a) has been very successful in modelling the development of birefringence with extension ratio (or equivalently draw ratio) in a rubber, and this is of a different shape from the predictions of the pseudo-affine deformation scheme (Eq. (30a)). There are also very significant differences between the predictions of the two schemes for P400- In particular, the development of P400 with extension ratio is much slower for the network model than for the pseudo-affine scheme. 

Gohlke H, Klebe G. Approaches to the description and prediction of the binding affinity of small-molecule ligands to macromolecular receptors. Angew Chem Int Edit 2002 41 2645-76. 

Bravi G, Wikel JH. Application of MS-WHIM descriptors 1. Introduction of new molecular surface properties and 2. Prediction of binding affinity data. 

---