176

■Bio-mathematics, Statistics and Nano-Technologies: Mosquito Control Strategies

9.3

THE IMPORTANCE OF CHEMOMETRIC MODELING IN DESIGN, CLAS-

SIFICATION AND SELECTION OF REPELLING COMPOUNDS

9.3.1

QSAR platform for modeling of repellent activity

The new repellent discovery and development process is very similar to the drug dis-

covery. The main dissimilarity in these two processes is financial side. In both of these

processes chemometric approach based on the quantitative structure-activity relationship

(QSAR) methodology stands as one of the crucial steps (Figure ) symbolically illustrates

the selection of target compounds through so-called Computer-Aided (Assisted) Drug De-

sign (CADD) funnel where initially huge number of compounds is significantly reduced to

several hits applying defined filters based on desirable molecular properties. QSAR mod-

eling correlates chemical structure with biological properties by using various linear and

non-linear methods (Natarajan et al. 2008).

9.3.2

Linear chemometric regression modeling of repellence index

Linear chemometric modeling is usually based on univariate linear regression (ULR),

multiple linear regression (MLR), principal component regression (PCR) and partial least

squares (PLS) regression. ULR is the simplest regression method that correlates an inde-

pendent variable with one dependent variable, while in MLR two or more independent

variables are correlated with one dependent variable. PCR is used when multicollinearity

among the independent variables is present so the utilization of MLR method is limited. In

this case it is necessary to normalize (scale) the variables. When multicollinearity occurs,

PLS regression method can also be used. This technique reduces the set of independent

Figure 9.2: Computer-aided drug design (CADD) with chemometrics and mathematical

modeling as crucial part of the process of selection of target or lead compounds and dis-

covery and development of new repellents (Created with Chemix (https://chemix.org)).