Virtual Screening in the Cloud Identifies Potent and Selective ROS1 Kinase Inhibitors

Area of Science:

• Medicinal Chemistry
• Computational Drug Discovery
• Oncology

Background:

• ROS1 rearrangements are implicated in 1-2% of non-small cell lung cancer (NSCLC) cases.
• Currently, no targeted therapies specifically designed for ROS1-positive NSCLC are available.
• TrkA serves as a relevant antitarget for selectivity in ROS1 inhibitor development.

Purpose of the Study:

• To identify potent and selective ROS1 inhibitors through virtual screening.
• To leverage shape-matching approaches for hit discovery in ultralarge chemical libraries.
• To develop novel therapeutic candidates for ROS1-driven cancers.

Main Methods:

• Ligand-based virtual screening using FastROCS technology on a 10^10 molecule subset of the AstraZeneca virtual library.
• Prioritization of libraries based on compound properties and medicinal chemistry assessment.
• Synthesis and biological evaluation of prioritized compounds, including docking to the ROS1 structure.

Main Results:

• Identification of several potent and selective ROS1 inhibitor series.
• Discovery of a lead compound with nanomolar potency against ROS1.
• Achieved over 1000-fold selectivity for ROS1 inhibition compared to TrkA.
• Preliminary structure-activity relationship (SAR) studies indicate optimization potential.

Conclusions:

• Conceptually simple shape-matching methods are effective for identifying potent and selective compounds from vast virtual libraries.
• This workflow demonstrates the utility and importance of virtual screening in early-stage drug discovery for targeted therapies.
• The identified ROS1 inhibitors represent promising leads for further development in NSCLC treatment.