Why BRAF-Δ, why now

Roughly half of all melanomas — and a meaningful slice of colorectal, thyroid, and lung cancers — carry a mutation in a gene called BRAF. The most famous version, BRAF V600E, has drugs against it. Patients live longer than they used to. That is a real, hard-won win.

But the tumor almost always comes back. Within a year or two, most cancers find a way around the drug. When we look at those resistant tumors under a sequencer, one of the escape routes that keeps showing up is a smaller, stranger sibling of V600E: a shortened form of the BRAF protein we call BRAF-Δ (BRAF-delta).

What BRAF-Δ actually is

Imagine BRAF as a switch that tells cells to grow. Normally the switch is spring-loaded — a whole section of the protein sits on top of it and keeps it off until the body says otherwise. BRAF-Δ has that safety section spliced out. The switch is stuck on. Worse, it clumps into pairs, and today's frontline drugs — the ones designed for V600E — bounce off those pairs. So the tumor grows again.

BRAF-Δ is not a fringe curiosity. It is one of the ways cancer wins the rematch.

Why it's still open

It would be tidy to say nobody has looked at BRAF-Δ. That isn't true — good academic groups have. But the drug industry has strong reasons to concentrate on the single mutations that hit the largest patient populations first. Resistance variants like BRAF-Δ sit in an awkward middle: too important to ignore, too niche to justify a $2B program on their own.

That awkward middle is exactly where a small lab can be useful. What we need:

• A clean, reproducible cell model that expresses BRAF-Δ at physiologic levels.
• An honest screen of existing and near-clinical inhibitors against that model.
• A short list — even three or four compounds — worth handing to a pharma partner or an academic clinician.

None of this requires a moonshot. It requires wet-lab discipline, a specific set of reagents, and about a year of focused work. That is the shape of our Phase-1.

Why now, and why us

Two things changed at the same time. First, protein-structure prediction and docking got good enough that a competent undergraduate with a laptop can generate real hypotheses about which compounds might fit a mutant kinase — hypotheses that used to require a whole computational-chemistry group. Second, CRO pricing on the basic assays we need has come down enough that a $100K program can actually buy data, not just buy meetings.

We are not claiming to be the only people who could do this. We are claiming that the work is worth doing, that the tools finally fit in a small lab's budget, and that a team of students with the right advisors is a reasonable place to put the next dollar.

If any of that resonates — as an investor, a partner, or a scientist who wants to poke holes in it — the door is open. Really open. We'd rather be wrong in public now than wrong in silence later.