Activity prediction models

K. Enslein, H. H. Borgstedt, M. E. Tomb, B. W. Blake, and J. B. Hart, Toxicol. Ind. Health, 3, 267 (1987). A Structure-Activity Prediction Model of Carcinogenicity Based on NCI/NTP Assays and Food Additives. 

Many of the products are biologically active. Predictive models of the toxicity of products, intermediates and by-products in the pure state and as mixtures would be useful at an early stage, and may help to decide between competing process options. 

Activity prediction is based on a list of models (i.e., QSAR models, pharmacophore models, etc.) that are maintained on the server. There is a second level of access so that only authorized users may be allowed to add or delete model entries. 

When enrichment episodes occur in the real world, but not in the laboratory under federal certification tests, real-world emissions are significantly higher than predicted. Further complicating emissions prediction is that aggressive driver behavior and complex traffic flow characteristics play a large role in enrichment occurrence. Current vehicle activity simulation models can predict average speeds and traffic volumes very well, but poorly predict the hard-accel-eration events that lead to enrichment. 

Mechanistic Approaches. Adequate and appropriate river-quality assessment must provide predictive information on the possible consequences of water and land development. This requires an understanding of the relevant cause and effect relationships and suitable data to develop predictive models for basin management. This understanding may be achieved through qualitative, semi-quantitative or quantitative approaches. When quantitative or semi-quantitative methods are not available the qualitative approach must be applied. Qualitative assessments involve knowledge of how basin activities may affect river quality. This requires the use of various descriptive methods. An example of this kind of assessment is laboratory evaluation of the extent to which increases in plant nutrients, temperature or flow may lead to accelerated eutrophication with consequent reduction of water quality. 

These pharmacophore techniques are different in format from the traditional pharmacophore definitions. They can not be easily visualized and mapped to the molecular structures rather, they are encoded as keys or topological/topographical descriptors. Nonetheless, they capture the same idea as the classic pharmacophore concept. Furthermore, this formalism is quite useful in building quantitative predictive models that can be used to classify and predict biological activities. 

Smith PA, Sorich MJ, McKinnon RA, Miners JO. Pharmacophore and quantitative structure-activity relationship modeling complementary approaches for the rationalization and prediction of UDP-glucuronosyltransferase 1A4 substrate selectivity. J Med Chem 2003 46 1617-26. 

Wang YW, Liu HX, Zhao CY, Liu HX, Cai ZW, Jiang GB. Quantitative structure-activity relationship models for prediction of the toxicity of polybrominated diphenyl ether congeners. Environ Sci Technol 2005 39 4961-6. 

Artursson P, Neuhoff S, Matsson P, Tavelin S (2006) Passive permeability and active transport models for the prediction of oral absorption. In Taylor JB, Triggle DJ (eds) Comprehensive medicinal chemistry II. Elsevier, Oxford, Sect 5.11... 

If an activity coefficient model is to be used at high pressure (Equation 4.27), then the vapor-phase fugacity coefficient can be predicted from Equation 4.47. However,... 

Although the methods developed here can be used to predict liquid-liquid equilibrium, the predictions will only be as good as the coefficients used in the activity coefficient model. Such predictions can be critical when designing liquid-liquid separation systems. When predicting liquid-liquid equilibrium, it is always better to use coefficients correlated from liquid-liquid equilibrium data, rather than coefficients based on the correlation of vapor-liquid equilibrium data. Equally well, when predicting vapor-liquid equilibrium, it is always better to use coefficients correlated to vapor-liquid equilibrium data, rather than coefficients based on the correlation of liquid-liquid equilibrium data. Also, when calculating liquid-liquid equilibrium with multicomponent systems, it is better to use multicomponent experimental data, rather than binary data. 

Prediction of liquid-liquid equilibrium also requires an activity coefficient model. The choice of models of liquid-liquid equilibrium is more restricted than that for vapor-liquid equilibrium, and predictions are particularly sensitive to the model parameters used. 

Inhibition of the hERG ion channel is firmly associated with cardiovascular toxicity in humans, and several drugs with this liability have been withdrawn. A number of studies show that basicity, lipophilicity, and the presence of aromatic rings [76] contribute to hERG binding. The 3D models of the hERG channel [77] are potentially useful to understand more subtle structure-activity relationships. In common with receptor promiscuity, both phospholipidosis and hERG inhibition are predominantly issues with lipophilic, basic compounds, and with the predictive models available, both risks should be well controlled. 

The model calculated in this manner predicts that two minerals, alunite [KA13(0H)6(S04)2] and anhydrite (CaSC>4), are supersaturated in the fluid at 175 °C, although neither mineral is observed in the district. This result is not surprising, given that the fluid s salinity exceeds the correlation limit for the activity coefficient model (Chapter 8). The observed composition in this case (Table 22.1), furthermore, actually represents the average of fluids from many inclusions and hence a mixture of hydrothermal fluids present over a range of time. As noted in Chapter 6, mixtures of fluids tend to be supersaturated, even if the individual fluids are not. 

The mutagenic activity of A-acyloxy-A-alkoxyamides reflects their interaction with the primary target, which in this case is bacterial DNA. The predictive model (Equation 3) allows discovery of structural factors that either increase or diminish DNA damage. Such effects can operate either upon binding to DNA or reactivity with DNA. Both types of structural impacts have been observed. 

The t value associated with each model descriptor, defined as the descriptor coefficient divided by its standard error, is a useful statistical metric. Descriptors with large f values are important in the predictive model and, as such, can be examined in order to gain some understanding of the nature of the property or activity of interest. It should be stated, however, that the converse is not necessarily true, and thus no conclusions can be drawn with respect to descriptors with small f values. 

Benigni, R., Andreoli, C., Conti, L., Tafani, P., Cotta-Ramusino, M., Carere, A., Crebelli, R. Quantitative structure-activity relationship models correctly predict the toxic and aneuploidizing properties of halogenated methanes in Aspergillus nidulans. Mutagenesis 1993, 8, 301-305. 

Gombar, V.K., Polli, J.W., Humphreys, J.E., Wring, S.A. and Serabjit-Singh, C. S. (2004) Predicting P-glycoprotein substrates by a quantitative structure-activity relationship model, fournal... 

For the purpose of this case study we will select Isopropyl alcohol as the crystallization solvent and assume that the NRTL-SAC solubility curve for Form A has been confirmed as reasonably accurate in the laboratory. If experimental solubility data is measured in IPA then it can be fitted to a more accurate (but non predictive) thermodynamic model such as NRTL or UNIQUAC at this point, taking care with analysis of the solid phase in equilibrium. As the activity coefficient model only relates to species in the liquid phase we can use the same model with each different set of AHm and Tm data to calculate the solubility of the other polymorphs of Cimetidine, as shown in Figure 21. True polymorphs only differ from each other in the solid phase and are otherwise chemically identical. 

Flexible optimal descriptors have been defined as specific modifications of adjacency matrix, by means of utilization of nonzero diagonal elements (Randic and Basak, 1999, 2001 Randic and Pompe, 2001a, b). These nonzero values of matrix elements change vertex degrees and consequently the values of molecular descriptors. As a rule, these modifications are aimed to change topological indices. The values of these diagonal elements must provide minimum standard error of estimation for predictive model (that is based on the flexible descriptor) of property/activity of interest. 

---