Machine learning driven drug repurposing strategy for identification of potential RET inhibitors against non-small...

Area of Science:

• Oncology
• Pharmacology
• Computational Chemistry

Background:

• Non-small cell lung cancer (NSCLC) is a leading cause of cancer death, with RET receptor activation driving a subset of cases.
• Current treatments for RET-driven NSCLC rely on multi-kinase inhibitors, highlighting the need for more specific therapeutic agents.
• Targeting the REarranged during Transfection (RET) pathway is crucial for developing effective NSCLC therapies.

Purpose of the Study:

• To identify novel, potent RET-specific inhibitors for RET-driven NSCLC through drug repurposing.
• To leverage high-throughput screening and machine learning for efficient compound identification.
• To validate potential drug candidates using in silico methods and assess their anticancer activity.

Main Methods:

• High-throughput screening of 2509 FDA-approved drugs from DrugBank.
• In-house machine learning model developed using BindingDB data for initial compound prediction.
• In silico validation including molecular docking, MMGBSA, and density functional theory analysis.
• Anticancer activity assessment of lead compounds against RET-expressing cell lines.

Main Results:

• A machine learning model predicted 497 (19%) compounds as potential actives.
• Two lead compounds, DB09313 and DB00471 (Montelukast), were identified through in silico validation.
• Montelukast demonstrated significant predicted anticancer activity against RET-driven NSCLC cells.
• Molecular simulations indicated Montelukast's potency as a RET inhibitor.

Conclusions:

• Montelukast is a promising drug candidate for the management of RET-driven NSCLC.
• Drug repurposing combined with computational methods can accelerate the discovery of targeted cancer therapies.
• Further investigation into Montelukast as a RET inhibitor for NSCLC is warranted.