Abstract.
Quantitative Structure-Activity Relationship (QSAR) study on 22 1H-indole-3-acetic acid derivatives with auxin activity was performed by means of Principal Component Analysis (PCA), Hierarchical Cluster Analysis (HCA), Partial Least Squares Regression (PLS) and Multiple Linear Regression (MLR). Molecular geometry of the auxins was optimized at MMFF94 and ab initio B3LYP 6-31G** levels. Modeling of complexes of some auxin molecules with the auxin binding protein 1 (ABP1) was also carried out. Parsimonius PLS and MLR models for prediction of optimal and half-optimal auxin concentrations for Avena L. Sativa coleoptile elongation were obtained with 15 auxins in the training set. HCA and PCA on data for the half-optimal concentration exhibit auxin clustering with respect to substitutent type and position, biological activity, and the size of the active site pockets of ABP1. Molecular graphics of ABP1 – NAA derivative complexes and of the coordination spheres around NAA (1-naphthalenic acid) hydrogen atoms in the ABP1 – NAA complex agrees well with the chemometrics/QSAR results.

Keywords.
1H-Indole-3-acetic Acid; Auxin Activity; Chemometrics; QSAR; Hierarchical Cluster Analysis; Principal Component Analysis; Partial Least Squares Regression; Substrate Modeling at the Active Site; Ab Initio; Molecular Mechanics.

Keywords Plus.