Is molecular alignment an indispensable requirement in the MIA-QSAR method?

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Cheminformatics

Background:

• Multivariate Image Analysis applied to Quantitative Structure-Activity Relationship (MIA-QSAR) has been effective for congeneric datasets.
• A key limitation of MIA-QSAR is its requirement for molecular scaffold alignment, restricting its use to structurally similar compounds.

Purpose of the Study:

• To introduce and evaluate the 2D-discrete Fourier transform (2D-DFT) for modeling structurally diverse, noncongruent chemical compounds.
• To expand the applicability of MIA-QSAR beyond congeneric datasets.

Main Methods:

• Application of 2D-discrete Fourier transform (2D-DFT) for image representation of noncongruent chemical compounds.
• Quantitative Structure-Activity Relationship (QSAR) modeling using the 2D-DFT features.
• Validation of regression models using statistical parameters.
• Comparative analysis against DRAGON molecular descriptors.

Main Results:

• The 2D-DFT MIA-QSAR approach successfully modeled a structurally diverse dataset of 100 compounds.
• The developed regression models demonstrated robustness and high predictive power.
• The 2D-DFT MIA-QSAR approach outperformed DRAGON molecular descriptors in predictive performance.

Conclusions:

• The 2D-DFT represents a significant advancement in MIA-QSAR, enabling the analysis of structurally diverse chemical libraries.
• This method overcomes previous limitations, opening new avenues for virtual screening and drug discovery.
• The superior performance suggests broad applicability for identifying novel molecular entities with desired therapeutic properties.