Peptide RX

Research OS demo flow

Idea -> fold -> candidate -> compare -> claim -> loop -> export.

One inspectable path through the Peptide RX system. Each step links to the live module that produced or displays the work, then resolves into a deterministic Claim Graph receipt.

Open Claim Graph Open Learning Loop View build receipts Launch pack
01 / Idea

Oral GLP-1 with reduced albumin binding

data-backed

Glucose-stimulated insulin secretion (GLP-1 / GIP wedge)

Open step ->
02 / Fold

CAMP WT

data-backed

Structure viewer, pLDDT, coverage, PAE context, and target biology.

Open step ->
03 / Candidate

PRX-GHSR-001

data-backed

Redacted sequence workup, gates, grade, selectivity, and provenance.

Open step ->
04 / Compare

Side-by-side structure comparison

data-backed

Compare folds, lanes, target context, and structural relationship.

Open step ->
05 / Timeline

studio_idea.created

data-backed

Latest event at 2026-05-05 08:07 UTC

Open step ->
06 / Evidence Claim

Initial digest — CAMP-LL37

data-backed

Grade C finding linked to the research graph.

Open step ->
07 / Claim Graph

Receipt graph for the public research spine

data-backed

Target, fold, AlphaFold metrics, candidate, finding, event, and export objects share one deterministic proof hash layer.

Open step ->
08 / Learning Loop

Feedback machine for proof debt

data-backed

Stored research objects become memory counts, agent scorecards, proof debt, and next-best actions.

Open step ->
09 / Export

Dossier-ready export surface

data-backed

Latest run: designing / 0 candidates.

Open step ->

what it proves

The system is more than a landing page.

Peptide RX now has a working path from research intent into structure context, candidate workups, comparison, event history, evidence-graded claims, and exports.

what stays honest

Evidence is not overclaimed.

Candidate sequences remain redacted on public pages, computational confidence is labeled as model evidence, and research-use limits stay visible across the flow.