A Multi-Objective Approach for Anti-Osteosarcoma Cancer Agents Discovery through Drug Repurposing

Area of Science:

• Oncology
• Computational Chemistry
• Drug Discovery

Background:

• Osteosarcoma is the most common primary malignant bone tumor.
• Current osteosarcoma therapies face limitations due to acute complications and late treatment effects, despite 5-year survival rates reaching 60-70%.

Purpose of the Study:

• To develop a multi-objective algorithm for repurposing drugs against osteosarcoma.
• To identify novel therapeutic agents for osteosarcoma by modeling molecules with known activity against relevant cell lines.

Main Methods:

• Utilized the ChEMBL database to model molecules active against HOS, MG63, SAOS2, and U2OS osteosarcoma cell lines.
• Developed predictive models with accuracy > 0.8, integrated into a multi-objective desirability function for virtual screening.
• Screened 2218 molecules from the DrugBank database using the validated multi-objective model.

Main Results:

• The final multi-objective model achieved BEDROC = 0.562, EF = 27.6, and AUC = 0.653 for the top 1% of screened molecules.
• Identified several top-ranked drugs including temsirolimus, paclitaxel, sirolimus, everolimus, and cabazitaxel, known antineoplastic agents.
• Discovered potential osteosarcoma treatments among broad-spectrum antibiotics and antiretroviral agents.

Conclusions:

• The developed algorithm effectively predicts repurposed drugs for osteosarcoma.
• Identified known antineoplastic drugs, antibiotics, and antiretrovirals warrant further investigation for osteosarcoma treatment.
• This approach offers a powerful strategy for discovering new chemotherapy regimens and treatment approaches for osteosarcoma.