Every medicine is approved on an average. Nobody is the average. The patient digital twin reads what the trials measured and asks one question for one person: will this settle my body, or stress it?
Medicine made by the body, for the body.
The dashboard in this film shows an earlier October build of the knowledge graph; the current counts are below. Read the article that goes with it →
About 7 in 100 drugs that enter human trials are approved (6.7%, 2014–2023). And some of the best-selling drugs in America help between 1 in 25 and 1 in 4 of the people who take them. Medicine built on the average fails most people twice: in the trial, and in the clinic.
6.7%of drugs entering Phase 1 reach approval (Citeline, 2014–2023). Biologics have historically done about twice as well.
Pick a patient. The trial results are real; the patients are illustrative.
Both are built on the same trial evidence and the same body-stress signal. They differ in who acts on what the twin sees.
Runs on the wearables people already own. It shows what the twin predicted next to what the body actually did, so a person can choose what to do about sleep, activity and the everyday factors that move their stress signal.
It never administers anything.
Works under a prescriber's care. The doctor sets the minimum and maximum dose, and the twin adjusts dosing only inside those limits as the body's signals change. Every decision is recorded.
It doses only inside the limits a prescriber has approved.
Registry results and published papers for the signalling proteins in our platform, scored on each trial's primary outcome against its control, on top of the SCKG, our knowledge graph of the published literature.
Bespoke treatments, matched to one body, built from the proteins the body already uses to heal.
Medicine made by the body, for the body.
Trial results from ClinicalTrials.gov and PubMed, reviewed 29 October 2026. Each trial is scored benefit, no difference or harm on its own primary outcome against its control. Side-effect figures are raw rates pooled across trials, not a statistical test. Approval rate: Citeline (Norstella), Biomedtracker data 2014–2023; BIO, Informa and QLS report 7.9% for 2011–2020. Biologics: BIO and Biomedtracker, 2003–2010. Responders: Schork, Nature 2015. PMID and node counts are distinct PMIDs and distinct nodes in the SCKG graph record of 28 October 2026.
Patients on this page are illustrative. This page shows the evidence the twin learns from. It is not medical advice.
Instructional Biology Inc. · www.instructionalbiology.com
Instructional Biology