Machine learning-based design, screening, and activity validation of topoisomerase I inhibitors

Area of Science:

• Medicinal Chemistry
• Computational Biology
• Oncology

Background:

• Topoisomerase I (TOP I) is crucial for genomic stability and proliferation.
• TOP I overexpression in cancers like lung, pancreatic, and breast malignancies makes it a key therapeutic target.
• Existing TOP I inhibitors suffer from poor stability, toxicity, and drug resistance.

Purpose of the Study:

• To develop a comprehensive QSAR framework for designing novel TOP I inhibitors.
• To overcome limitations of traditional QSAR methods using diverse molecular fingerprints and machine learning algorithms.
• To identify and validate potent TOP I inhibitors for targeted cancer therapy.

Main Methods:

• Systematic screening of 550 high-activity compounds from multiple databases.
• Development of thirty QSAR models using five molecular fingerprint types and six machine learning algorithms.
• Rational design of candidate inhibitors using an optimized artificial neural network and SAFE methodology, followed by activity prediction, drug-likeness, and ADMET profiling.

Main Results:

• Seven promising candidate TOP I inhibitors were identified through computational evaluation.
• Three compounds demonstrated cytotoxicity via MTT assays.
• Four novel compounds underwent further characterization using molecular docking and dynamics simulations.

Conclusions:

• The integrative QSAR approach provides a robust foundation for designing and optimizing TOP I inhibitors.
• This research facilitates the development of targeted therapies against TOP I-associated cancers.
• The identified compounds show potential for overcoming current therapeutic challenges.