Marine keys for Immune locks
An AI-driven quest to discover novel marine compounds targeting cancer immunotherapy.
Introduction
The field of immuno-oncology seeks to harness the body’s immune system to fight cancer. While checkpoint inhibitors targeting proteins like PD−1 have been revolutionary, the need for novel therapeutic targets with untapped mechanisms is critical. Orphan nuclear receptors, whose natural activating molecules (endogenous ligands) are unknown, represent a frontier in drug discovery. This project focuses on NR2F6, an orphan receptor identified as a key regulator in both cancer progression and immune cell function, making it a high-value, dual-action target.
Marine natural products, particularly alkaloids from species like sponges, offer immense, underexplored chemical diversity, providing a rich source for discovering first-in-class molecular structures. By combining a novel biological target with a diverse chemical library, this project aims to bridge a critical gap in modern drug development.
Hypothesis
Novel marine alkaloids, sourced from chemically diverse and understudied marine species, will function as allosteric modulators or direct agonists/antagonists of the orphan nuclear receptor NR2F6. The binding of these compounds to NR2F6 will alter its regulatory activity, leading to an enhancement of anti-tumor immunity, providing a new therapeutic avenue for cancer treatment.
Project Summary
This project is a computational drug discovery campaign executed in a rapid, three-phase sprint.
Data Curation:
First, the 3D structure of the NR2F6 protein is obtained and prepared for analysis. Simultaneously, a virtual library of 50-100 marine alkaloids is compiled by data mining chemical databases like PubChem and ChEMBL, and their structures are prepared for screening.
Virtual Screening:
Using a hybrid cloud platform of Google Colab and AWS, the marine alkaloid library is computationally “docked” against the binding pocket of NR2F6 using software like AutoDock Vina. This process calculates and ranks each molecule’s binding affinity, identifying the most promising candidates.
Hit Analysis:
The top 10 candidates are then analyzed for drug-like properties (ADMET) and potential toxicity. The final deliverable is a lead compound with a predicted binding mode, visualized as an interactive 3D model suitable for a portfolio.