Improvement of Multivariate Image Analysis Applied to Quantitative Structure–Activity Relationship (QSAR) Analysis by Using Wavelet‐Principal Component Analysis Ranking Variable Selection and Least‐Squares Support Vector Machine Regression: QSAR Study of Checkpoint Kinase WEE1 Inhibitors

Abstract

Inhibition of tyrosine kinase enzyme WEE1 is an important step for the treatment of cancer. The bioactivities of a series of WEE1 inhibitors have been previously modeled through comparative molecular field analyses (CoMFA and CoMSIA), but a two-dimensional image-based quantitative structure-activity relationship approach has shown to be highly predictive for other compound classes. This method, called multivariate image analysis applied to quantitative structure-activity relationship, was applied here to derive quantitative structure-activity relationship models. Whilst the well-known bilinear and multilinear partial least squares regressions (PLS and N-PLS, respectively) correlated multivariate image analysis descriptors with the corresponding dependent variables only reasonably well, the use of wavelet and principal component ranking as variable selection methods, together with least-squares support vector machine, improved significantly the prediction statistics. These recently implemented mathematical tools, particularly novel in quantitative structure-activity relationship studies, represent an important advance for the development of more predictive quantitative structure-activity relationship models and, consequently, new drugs.

MIA–QSAR coupled to principal component analysis-adaptive neuro-fuzzy inference systems (PCA–ANFIS) for the modeling of the anti-HIV reverse transcriptase activities of TIBO derivatives

“…In line with this, many descriptors usually do not explain considerable variance in Y because most regression methods, such as PLS and N-PLS, do not account for non-linearities. A recent study on checkpoint kinase WEE1 inhibitors illustrates well the importance of considering nonlinearity during the regression step in a QSAR modeling [20].…”

“…Multivariate image analysis (MIA) applied to QSAR has provided predictive models for several compound classes [10][11][12][13][14][15][16][17][18][19][20][21], and was proved to be a valuable tool in proposing new active entities. In the MIA-QSAR method, descriptors are pixels (binaries) of 2D images -chemical structures.…”

“…In the MIA-QSAR method, descriptors are pixels (binaries) of 2D images -chemical structures. In addition to the widely used PLS regression, some mathematical approaches have been coupled to MIA-QSAR in order to achieve progressively more predictive models, such as N-way and non-linear methods [18][19][20][21][22]. In this work, principal component analysis-adaptive neuro-fuzzy inference systems (PCA-ANFIS) technique was applied together with MIA-QSAR to verify the enhancement in predictive ability of the bioactivities of TIBO derivatives when compared to previous works [4][5][6][7][8][9] and to MIA-QSAR models based on well established regression methods.…”

aug-MIA-QSPR study of guanidine derivative sweeteners

“…In line with this, a quantitative structure-activity relationship (QSAR) method based on multivariate image analysis (MIA), named MIA-QSAR, has shown to be accurate in estimating bioactivities of a variety of drug-like compounds [15][16][17][18][19][20][21][22]; also, it has been successfully used to model other physical properties, such as NMR chemical shifts [23], electrophoretic profiles [24] and boiling points [25]. This method is based on the relationship of images (pixels, numerically described as binaries), which are chemical structures built using software for chemical drawing, with the respective dependent variables (physical, chemical or biological properties).…”

Feature selection method based on fuzzy entropy for regression in QSAR studies

“…But it should be noted that in previous work we carried out a QSAR study of this class of compounds, but using MIA-QSAR, which was compared with CoMFA and CoMSIA [15]. The main aim of this work was to introduce new feature selection in the QSPR field that could be used to predict the activity of tyrosine kinase enzyme WEE1 derivatives from their molecular structure alone using multiple linear regressions as a fast and simple regression method.…”