Linear Regression QSAR Models for Polo-Like Kinase-1 Inhibitors

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Polo-like kinase-1 (PLK1) is a key regulator of cell division, making it a promising target for cancer therapy.
• Developing effective PLK1 inhibitors requires understanding the relationship between chemical structure and biological activity.
• Existing quantitative structure-activity relationship (QSAR) models for PLK1 inhibitors can be complex.

Purpose of the Study:

• To develop a novel, conformation-independent QSAR model for a diverse set of PLK1 inhibitors.
• To identify key molecular descriptors that significantly influence PLK1 inhibitor bioactivity.
• To provide a simpler and improved QSAR model compared to previously reported methods.

Main Methods:

• Compilation of a dataset of 530 PLK1 inhibitors from the ChEMBL database.
• Generation of 26,761 molecular descriptors using freeware tools (PaDEL, Mold², QuBiLs-MAS).
• Application of multivariable linear regression with a variable subset selection technique and balanced subset partitioning for model development and validation.

Main Results:

• Identification of a robust linear QSAR model that effectively correlates structural features with PLK1 inhibitor bioactivity.
• The model utilizes a selected subset of molecular descriptors, simplifying the structure-activity relationship.
• The developed QSAR model demonstrates improved performance and simplicity over existing models.

Conclusions:

• The proposed linear QSAR model offers a valuable tool for the design and optimization of novel PLK1 inhibitors.
• This study highlights the utility of diverse descriptor sets and robust modeling techniques in drug discovery.
• The findings contribute to a better understanding of the structural requirements for potent PLK1 inhibition.