Asclepius
── /methodology ──

Methodology

The framework's intellectual foundation. Seven writeups, every empirical claim cited.

Each writeup is also a standalone document; cross-links between them stay within the methodology folder. A senior reader can land on any file and follow the citations from there.

Read this first

Product thesis — what Asclepius is, what it's becoming, what it's a proof of

Applies The Loom's six-phase product framework (Sense / Envision / Seed / Surface / Evolve / Synthesize) to Asclepius itself. Reframes the project from 'portfolio biotech-valuation tool' to 'first proof point of a method for productizing skill-based AI methodology.'

Read the product thesis →
Technical writeups

Probability of Success — the full framework reference

The end-to-end documentation of how Asclepius computes cumulative likelihood of approval. Every multiplier in the chain is cited; every magnitude is grounded in either a peer-reviewed paper or a primary industry source; deviations from the published literature are explicit.

6 primary sources cited

Reflexivity-adjusted Probability of Success

Sponsor capital position structurally adjusts a clinical asset's probability of success. This is the headline methodology choice in Asclepius and the differentiator versus textbook rNPV.

4 primary sources cited

Risk-adjusted NPV — phase-gated cash flows, Monte Carlo priors, and discount conventions

The full mechanics of how Asclepius computes risk-adjusted NPV. Documents the phase-gated cash-flow structure, the four Monte Carlo prior distributions, the mid-year discount convention, and the closed-form-vs-simulation relationship. Every methodology choice is cited or labeled as analytical convention.

5 primary sources cited

8-pillar diligence scorecard — the qualitative overlay on the quantitative core

Documents the 8 pillars Asclepius uses to convert qualitative diligence judgments into a single weighted score and a recommendation tier. Seven pillars are practitioner-standard biotech BD scorecard items; the 8th (computational infrastructure) is the framework's novel contribution. Documents the red-flag auto-cap and green-flag boost mechanics, and consolidates the methodology's Known Limitations.

4 primary sources cited

Worked example — Adagrasib retrospective backtest

Applies the Asclepius framework to adagrasib (Mirati Therapeutics) using only public information available before the KRYSTAL-12 Phase 3 readout (June 2022 cutoff). Framework outputs reconstruct the BMS $4.8B acquisition price (October 2022) within ~5%.

6 primary sources cited

Signaling equilibrium — the formal foundation of the reflexivity adjustment

Derives the reflexivity-adjusted PoS framework from Spence (1973) signaling theory. Maps the costly-signal mechanism onto clinical trial design and cites the modern empirical evidence that strategic-disclosure behavior is operative in the pharmaceutical industry.

7 primary sources cited

07-bayesian-persuasion-disclosure

08-calibration-dashboard

09-ml-pos-prior

10-public-prediction-log

11-portfolio-sizing

How to read the methodology folder