Chemical indices and methods of multivariate statistics as a tool for odor classification

Mammalian olfactory receptors: molecular mechanisms of odorant detection, 3D-modeling, and structure-activity relationships

This chapter describes the main characteristics of olfactory receptor (OR) genes of vertebrates, including generation of this large multigenic family and pseudogenization. OR genes are compared in relation to evolution and among species. OR gene structure and selection of a given gene for expression in an olfactory sensory neuron (OSN) are tackled. The specificities of OR proteins, their expression, and their function are presented. The expression of OR proteins in locations other than the nasal cavity is regulated by different mechanisms, and ORs display various additional functions. A conventional olfactory signal transduction cascade is observed in OSNs, but individual ORs can also mediate different signaling pathways, through the involvement of other molecular partners and depending on the odorant ligand encountered. ORs are engaged in constitutive dimers. Ligand binding induces conformational changes in the ORs that regulate their level of activity depending on odorant dose. When present, odorant binding proteins induce an allosteric modulation of OR activity. Since no 3D structure of an OR has been yet resolved, modeling has to be performed using the closest G-protein-coupled receptor 3D structures available, to facilitate virtual ligand screening using the models. The study of odorant binding modes and affinities may infer best-bet OR ligands, to be subsequently checked experimentally. The relationship between spatial and steric features of odorants and their activity in terms of perceived odor quality are also fields of research that development of computing tools may enhance.

A semiparametric statistical approach for forecasting SO₂ and NOx concentrations

This study develops a new semiparametric statistical approach for urban air quality forecasting. Compared to conventional approaches, the semiparametric approach allows the model users to benefit from the positive aspects and alleviate the negative ones of parametric and nonparametric approaches. Two advantages of the approach lie in (1) the interpretation of the data set being easily decoded and used by the model and (2) its capability in dependence on prior assumption. To illustrate the performance of the proposed approach, three semiparametric regression models (i.e., linear-, quadratic-, and interactive-based semiparametric regression) are applied to an air quality forecasting problem in the city of Xiamen, China, and satisfactory training and prediction performance are obtained. The three models are also compared to three parametric and two nonparametric regression models. The results indicate that the predictive accuracy of semiparametric regression models is higher than those obtained from the parametric and stepwise cluster analysis models. However, the proposed three semiparametric regression models could be much favored, since they can be achieved more easily and rapidly than the artificial neural network model.