Log — Life Sciences Wiki
Append-only history. Entries start ## [YYYY-MM-DD] <op> | <title>, where <op> is
ingest, query, quality or split.
[2026-09-17] split | life-sciences-wiki created from _inbox (3 sources)
Curator named the spoke on Telegram (msg 2147) after the third bio-ai park. Ingested:
hippocampus-aging-epigenome (T1, via ScienceDaily/Technology Networks; primary 403),
autism-protein-interaction-map (T2, CNN; primary unread), israel-ai-bio-consortium (T3, CTech,
re-read today). Concepts biological-foundation-model, single-cell-multiomics; people bing-ren,
nevan-krogan. israel-innovation-authority reused cross-wiki from ai-governance-wiki. Inbox sweep:
no other record fits the domain.
[2026-09-27] ingest | Two parallel neural ectoderm progenitors contribute to the developing brain (Stanford Medicine)
Routed by the hub from med.stanford.edu/news/all-news/2026/09/two-separate-brains.html (runner-up: psychology-wiki, noted as cross-spoke adjacency in synthesis — the substance here is structural/ developmental, not cognitive/behavioral). Wrote two-brain-progenitor-lineages (T1, Nature Neuroscience, 18 Sep 2026; journal page login-gated, held via the Stanford press release — same reading-limit pattern as hippocampus-aging-epigenome). Finding: forebrain/midbrain and hindbrain develop from two non-overlapping progenitor lineages (Otx2 vs. Gbx2) fixed apart at the chromatin level, conserved ~550 Myr across chicken/zebrafish/acorn worm; first working human hindbrain motor neurons from stem cells. Entities (3 of 8 budget): kyle-loh (senior author, url verified against profiles.stanford.edu), carolyn-dundes (co-first author, url verified against vpge.stanford.edu), rayyan-jokhai (co-first author, no verifiable profile URL found — left off rather than guessed). stanford-university reused cross-wiki from agentic-tooling-wiki, not duplicated. Gap-relevance: doesn't close a named growth edge; opens a new thread ("where brain regions come from") recorded in synthesis, with one evidenced but unconfirmed link to the spoke's existing drug-discovery thread (hindbrain motor neurons as an SMA/ALS disease-modeling substrate).
[2026-09-22] quality | research pass — edges 1, 2 and 4
Via the hub quality cycle's research pass (spoke never hunted before). Edge 1 closed: geneformer (T1, Nature 2023, read from PMC) — the first held biological foundation model, with a wet-lab-confirmed prediction. Beside it, single-cell-foundation-model-zero-shot (T1, Genome Biology 2025), an outside zero-shot test in which a gene filter beats it; recorded as a tension in synthesis. Edge 4 half-closed: rentosertib-phase-2a (T1 for the trial facts, efficacy flagged as sponsor-reported); residue written as edge 5, and edge 6 added for the zero-shot question. Edge 2 half: the hippocampus abstract (PubMed) now read — "resembling" monocyte-derived microglia, not traced from bone marrow. Entity: insilico-medicine (Wikipedia + Wikidata verified). biological-foundation-model rewritten around the held instance.
[2026-10-11] research | AI-derived drug followed from outside the company (edge 45) — not closed, status changed
Research pass hunt. Rentosertib's phase 3 (GENESIS-IPF-3, NCT07687459) started 2026-09-09: 320 patients, 52 weeks, primary completion estimated 2029-10-30. Insilico is sole sponsor, so the edge stays open. Added to rentosertib-phase-2a from the ClinicalTrials.gov API record.