Cancer first, then the diseases that inherit the lessons
Starting in cancer is where selectivity, safety, and pathway response can be learned together. Neurodegeneration is how we find out whether those lessons survive a different organ.
Starting with cancer is not a branding decision. It is a place where the system can learn several hard things at once: target engagement, whether diseased cells die and healthy ones do not, what the transcriptional aftermath looks like, and which safety liabilities show up early. Those are not cancer-only skills. They are the skills you need the moment you point the same machinery at a neuron.
What actually transfers
A model does not "transfer to Parkinson's" because a slide says so. What transfers, if you are lucky and careful, is a set of habits.
- How to rank binding when the target is awkward, including RNA rather than a comfortable pocket
- How to treat selectivity as a difference between two cellular contexts, not a single potency
- How to demand CNS exposure and a clean safety profile before calling something a candidate
- How to refuse a prediction that sits outside the applicability domain
Neurodegeneration is a stern test of all four. The tissue is harder to reach, the therapeutic window is less forgiving, and a toxicity you might tolerate in an oncology cytotoxin is disqualifying in a chronic CNS drug.
The programs that make this concrete
The neurodegenerative thread is already in the collaborations, not in a future roadmap paragraph. LRRK2 inhibitors for Parkinson's, with WEHI, are being designed for blood-brain barrier exposure and for safety rather than for a potency screenshot. C9orf72 RNA-targeted molecules for ALS and frontotemporal dementia, with McGill's The Neuro, are a direct bet that Gnosis I RNA can reach biology protein-pocket drugs do not cover. TRPM2 inhibitors with Griffith sit on oxidative-stress injury in neurons, which is the unglamorous middle of several neurodegenerative stories.
Cancer remains the proving ground, because that is where selectivity, cytotoxicity, and pathway response can be learned against data that actually exists. The neurodegenerative programs are how we find out whether those lessons survive contact with a different organ.
The ambition is not to collect diseases. It is to learn a way of choosing molecules that still works when the disease changes.
Vecentra Research
Vecentra outputs are intended for research use only and are not validated for clinical diagnosis or therapeutic decision-making.