Magellan AI Identifies and Optimizes Molecules Beyond the Public Domain

In an interview with Pharmaceutical Executive, Gene Mack, CEO, Gain Therapeutics, shared promising updates on the company’s lead candidate, GT-02287, a novel therapy in development for Parkinson’s disease. Enrollment for the Phase Ib open-label trial wrapped three months ahead of schedule, driven by growing clinician confidence in the drug’s safety and therapeutic potential. Originally designed for patients with a GBA1 gene mutation—a known disruptor of the GCase enzyme—GT-02287 has shown early signs of benefit across a broader Parkinson’s population. With biomarker data expected in Q4 2025, Gain is preparing for a pivotal Phase II trial in early 2026 while weighing financing options and potential strategic partnerships, all hinging on forthcoming efficacy and biomarker insights.

Pharmaceutical Executive: Can you share more about the Magellan AI platform’s capabilities, and how it’s being used to identify or optimize molecules like GT-02287?

Gene Mack: Magellan is the result of a lot of foundational work built on top of a computational chemistry platform. Initially, we had a screening tool that allowed us to evaluate known molecules against specific protein targets. These target proteins are the focus of our small molecule drug development efforts.

What makes Magellan different is its ability to go further. It starts by analyzing a protein structure and identifying novel, potentially important binding sites. From there, it calculates the binding kinetics to predict which molecules—whether known or not yet synthesized—might fit into that binding pocket.

This opens up drug discovery well beyond the molecules already available in the public domain. Magellan allows us to explore new chemical space by designing molecules that don’t yet exist but could theoretically be synthesized. Once the platform identifies a promising candidate, it’s up to us to synthesize it, assess its stability, and determine whether it’s feasible to produce.

Ultimately, the goal of Magellan is to expand drug development opportunities by using advanced physical and chemical modeling to go beyond what’s currently known.

Full Interview Summary: Enrollment for the Phase Ib trial of GT-02287, a novel therapeutic candidate for Parkinson’s disease, was completed three months ahead of schedule. While the trial targeted a small cohort, momentum increased as clinicians grew more comfortable with the drug’s safety profile and potential. Initially cautious due to the investigational nature of the therapy, sites became more confident in recommending the study, reflecting growing optimism about GT-02287’s promise.

The ongoing Phase Ib trial is an open-label pilot involving 15 to 20 patients with Parkinson’s disease, including those with idiopathic PD and those with a GBA1 gene mutation. This gene mutation disrupts the production of glucocerebrosidase (GCase), a key enzyme that GT-02287 is designed to restore. Although the drug was initially developed for GBA1-mutated patients, early findings suggest it may benefit a broader PD population. The Phase Ib readout, expected in Q4 2025, will include biomarker data from cerebrospinal fluid to assess differential efficacy between patient subtypes.

Looking ahead, a larger, double-blind Phase II trial is planned for early 2026, enrolling 100–200 patients to validate efficacy signals in a blinded setting and mitigate bias seen in open-label designs.

GT-02287 was developed using the company’s Magellan AI platform, which identifies novel binding pockets on target proteins and screens both known and hypothetical molecules. Magellan extends drug discovery into previously unexplored chemical space by predicting interactions that can be synthesized and tested.

Partnership discussions are ongoing with large pharma, biotech, and mid-cap firms. Final decisions around financing or partnering will depend on Phase Ib results. The company is prepared to either self-fund the $50–60M Phase II trial or advance through a strategic collaboration, depending on the biomarker and efficacy data in Parkinson’s patients.