← World Globe Sickbay Porch Strain Catalog Methods Paper

Oncology · Sickbay Cathedral Porch

world.hitme.dev · living instrument · helpers not doctors · in-silico only

Teaching note. Sickbay is a window that prints a bench padd when someone asks for help — not a hospital claim. We teach assays, spores, and kill criteria. Labels (basin, attractor, P-gp) are not lemmas until a theorem-preserving homomorphism is proven on measured silicon or wet-lab receipts.

Brian Philip King · ORCID 0009-0003-2718-2676 · Forge HM1 · py=0 · George + Kids hands · 2026-09-15

HELPERS NOT DOCTORS WET=0 · BENCH NOT SENT 3 OPENBIND SPORES SEEDED 256 COIL LEDGERS

1. Sickbay Bench Padd · TNBC Paclitaxel Resistance

SICKBAY CATHEDRAL PADD — plain work order under the glass globe.

Hypothesis: C_raft ↓ ⇒ J_P-gp ↓ ⇒ F(t=1h) ↑ ⇒ IC50(231-PacR) ↓

Host gate: MCF-10A viability ≥ 80% at tumor IC50 exposure.

Lines: MDA-MB-231-PacR · MDA-MB-231-P · MCF-10A (host gate)
Plot 1: Rhodamine-123 / Calcein-AM efflux surrogate (fold vs vehicle)
Plot 2: 100 nM paclitaxel · 1.0 h · LC-MS/MS intracellular peak area
Plot 3: 72-h IC50 · 8-point dilution 0.1–1000 nM · MCF-10A gate

CRO methods sheet Plate map TSV ECL protocol outline

Does not assert clinical outcome. Assay map only — not sent to labs (track 2 = 0%).

2. Nursery Spores · OpenBind EV-A71 (structural bench)

Three vspace spores seeded from OpenBind metadata inventory (699 structures · 649 pKD rows · py=0 count only · ML notebook not run).

Spore	Group	experimental_pKD	Law
spore_23558_openbind_x7259a	x7259a	7.938	inventory label
spore_23559_openbind_x7135a	x7135a	7.760	inventory label
spore_23560_openbind_x6832a	x6832a	7.714	inventory label

pKD values are catalog metadata, not wet-lab replication on Forge. Coils expanded 32→256 · cradles=24 · babies=14 · receipt FORGE_NURSERY_EXPAND_OPENBIND_RECEIPT_NOW on door :1700.

73-strain public catalog →

3. Oncology Methods Audit (historical · retired headline)

RETIRED (2026-09-09): 91.26% / 86.11% were in-sample METABRIC rule recovery — not external PAM50 validation. Do not cite as performance.

METABRIC N=1,980	Linear OK	Linear fail	Total
Discrete OK	1,674	133	1,807
Discrete fail	31	142	173

Independent PAM50 check: 49.37% full cohort · 55.69% 5-class subset. Future claims need external fold-wise Δ above majority baseline.

Figshare DOI (erratum pending) Full audit on gotokell porch →

4. NCI ITCR Grant Frame · Methods Not Hope

One-page frame minted on Forge door :1700 · helpers not doctors · tools in the zip · no cure · no 91.26 headline.

OID=oid.forge.itcr.one_page_frame.20260915T134500Z · stamp=FORGE_NCI_ITCR_ONE_PAGE_FRAME_NOW.txt

Aim 1: Ship atom_emt_reversibility + confound_audit as one-command pack (W1B GSE147405 self-test).
Aim 2: Pre-reg TCGA-BRCA + GSE96058 replication · seal-before-values · BLOCKED until rows pulled.
Aim 3: Independent Mac/Linux dry-run returns receipt stub only · not a seminar.

Mechanism: NCI ITCR R21/U01 class · open-source awk atoms · py=0 · Devin 5× audit before external send.

NIH ITCR program →

5. Explicit limits

Not a diagnostic · not a therapy · not FDA cleared.
Basin / Markov language = discrete audit labels unless externally validated.
OpenBind spores = structural nursery pots · not drug approval evidence.
Capillary latch WET=0 on Forge — money lane HOLD · structural porch only.