First priors publish in late October 2026

NeuroPriors

A neurologist's priors on neuroimmunology and CNS genetic medicines, dated before the data.

Before an FDA decision or a trial readout, I publish one number: the probability that the event goes the way the question asks. With it come the reasoning, the bar the data must clear, and what would change my mind. When the event happens, the prior is scored in public, and the misses are listed first.

Daniel Branco, MD, PhD, MBA Neurologist · Powered by BioAlma

What a prior contains

Each prior covers one scheduled public event: an FDA decision, a trial readout, a regulatory meeting or a deal. It is written in this order.

The event. What, who and when, linked to the company's own guidance or filing.
The question. One YES or NO question, with its resolution source, delay rule and void rule, all fixed in advance.
The base rate. How often events of this kind have gone YES, with its source.
The adjustments. What places this case above or below the base rate.
The number. One probability, and the date it was published.
The bar. For a readout, the minimum result that counts as success and what a strong result looks like.
What would change my mind. Two or three observations that would move the number.
If YES, if NO. What each outcome means for the field, for partnering and for financing.

How the record is kept

The rules exist so that anyone can check the record without taking my word for it.

The first number counts

The number published first is the one scored. Amendments are dated, shown under the original and scored separately. They never replace it.

Misses first

Resolved misses are listed before any hits, each with what was right, what was wrong and the lesson.

Timestamped

Each prior is archived on the Wayback Machine the day it is published, and the SHA-256 hash of its text is published with it. Any later edit is visible.

Calibration

Resolved priors are grouped by stated probability and compared with how often YES happened. The chart becomes meaningful at about 30 resolved priors.

Rules before results

Every question states in advance what happens if the date moves, and when it is voided instead of scored. Voided priors stay on the page with the reason.

One record

Every prior counts toward one scorecard, across all lanes. Lane results are shown once a lane has enough resolved priors to mean something.

Independence

Public information only. No material non-public information is used.
No positions in covered companies, and no payment from any company for a prior.
No priors on companies that are clients of BioAlma or in active talks with it. Client names never appear.
These are scientific, clinical and regulatory judgments. They are not investment advice.

Coverage

Public companies and public catalysts in two lanes.

Neuroimmunology

Multiple sclerosis, myasthenia gravis, NMOSD, MOGAD, CIDP, autoimmune encephalitis and stiff person syndrome, including CAR-T, cell therapies and T-cell engagers in these diseases.

CNS genetic medicines

Antisense, RNA, gene and cell therapies for genetic neurological and neuromuscular disease, and precision therapies for the genetic epilepsies.

Cadence: one prior a week, and one longer memo a month.

Who writes it

Daniel Branco is a neurologist and neuroscientist. He began his career mapping the living human brain for epilepsy surgery, then worked in healthcare private equity and venture diligence and as a hospital chief medical officer.

Until 2026 he was Senior Director of Corporate Development and Strategy at Artiva Biotherapeutics, a clinical-stage cell therapy company, responsible for financial planning through its IPO and then for medical and competitive intelligence.

He founded BioAlma, the intelligence platform that prepares the research behind each prior. The judgment and the number are his.

The first priors publish here in late October 2026, and the scorecard opens with them. Until then, the weekly CoCreating.Bio newsletter is the place to follow along.

Read the newsletter