STRC Research

STRC Morning Report 2026-04-28

9 min read

STRC Morning Report — 2026-04-28

Synthesis-only. First morning report since 2026-04-25; recaps overdue verdicts from 2026-04-26 and 2026-04-27 plus this morning’s lit-watch deliveries. No compute, scripts, hubs, or proofs touched this run.

TL;DR

  1. h09 promoted tentative-S → S on 2026-04-27 via Phase 4m Kd soft-floor PASS (R²=0.690, predicted Kd_F = 1.48 mM closes Phase 4j conservative gate). Reviewer pack now staged.
  2. h01 v5.3 sulfonamide kinetic-selectivity falsification statistically robust at n=20 (Phase 5q τRAMD 2026-04-26): σ_within = 1.358× mean / 1.167× median; STRC:TRPM4 lemma bound 2.72× / 2.33× ≪ 5× gate. Wetlab handoff packet v0.1 filed 2026-04-27 (compound sheet + gap matrix + RFQ template).
  3. Top open question: h01 hub excerpt claims “deliv 3→4” from Phase 8h-lite RWM evidence but deliv: frontmatter still reads 3 — silent re-tier inconsistency, needs Egor decision before next ranking edit.

Ranking state (before → after)

No tier moves this run (synthesis only). Recap of the moves that landed since last morning report:

#hub2026-04-252026-04-28trigger
h01A · 4 / 3 / 4A · 4 / 3 / 4A · 4 / 3 / 4Phase 5q n=20, Phase 8h-lite, Phase 9b smoke, wetlab handoff — A held; excerpt-vs-frontmatter mismatch on deliv (see Sanity check)
h03S · 5 / 5 / 4S · 5 / 5 / 4S · 5 / 5 / 4unchanged — awaiting Ultra-Mini gBlock order
h09tentative-S · 5 / 4 / 3S · 5 / 4 / 3S · 5 / 4 / 3Phase 4m PASS promoted on 2026-04-27 15:22; Phase 4k inconclusive (geometric yes, confidence-supported no) didn’t reverse; Phase 4p Boltz-2 supportive

All other hubs (h02 / h04 / h05 / h06 / h07 / h08 / h10 / h11 / h12 / h13 / h14 / h15 / h16 / h26 / h27) unchanged.

What shipped since 2026-04-25 (recap, not this run)

h01 — pharmacochaperone (A held)

  • Phase 5q v5.3 within-target τRAMD n=20, 2026-04-26 21:37 UTC. 60/60 UNBOUND in 10.8–59.8 ps biased MD on Phase-5d-truncated 151_e1659a fragment; ~50 min M5 Max CPU OpenMM. σ_within: 1.45× (n=5 pilot) → 1.358× (n=20 mean) / 1.167× (n=20 median). Lemma cross-target STRC:TRPM4 bound = 2.72× / 2.33× ≪ 5× kinetic-selectivity gate. Rank-2 flipped vs n=5 (adamantyl_-Cl > 1-indanyl_-Cl); affinity-rank-1 ≠ τ-rank-1 confirmed real. → STRC h01 Phase 5q v5.3 STRC Within-Target tauRAMD 2026-04-26
  • Phase 8h-lite, 2026-04-26. Five light proofs closing paper-claim gaps: K1141 pocket V=1145 ų, phobic 0.61, V_pocket/V_lig 2.75 (Halgren druggable); RWM permeability 7.6× TMPA → ~60% basal-turn applied conc at 90 min; Vina pose ensemble-stable (K1141 NZ SD 1.1 Å, CONHO⁻↔NZ 5.0±0.6 Å across 20 Phase 5d snapshots); Tanimoto vs 12-compound ototoxin panel max 0.127 (aspirin), no class motif. Hub excerpt logs deliv claim 3→4 from RWM result but frontmatter unchanged — flagged below.
  • Phase 9b geometry sanity + repaired matched Amber + min-relax multi-frame MM-GBSA, 2026-04-27. WT ΔG_GB = −15.84 ± 0.46 SEM kcal/mol; E1659A = −18.18 ± 0.09; Δ(E1659A − WT) = −2.35 kcal/mol. Caveated robustness smoke; production explicit-solvent holo MD/MM-PBSA still required for paper-grade. → STRC h01 Phase 9b Matched Amber Input Compatibility 2026-04-27 · STRC h01 Phase 9b Min-Relax Multi-Frame MM-GBSA Smoke 2026-04-27
  • Wetlab handoff packet v0.1, 2026-04-27. compound_sheet.csv + gap_matrix.csv + rfq_email.md at ~/STRC/hypotheses/h01-pharmacochaperone/artifacts/wetlab_handoff_2026-04-27/. Reframes h01 as target-engagement feasibility, not lead-candidate nomination. → STRC h01 Wetlab Handoff Packet 2026-04-27

h09 — synthetic peptide hydrogel HTC (tentative-S → S)

Vault-wide infrastructure (2026-04-26 / 27)

New literature relevant to current S/A tier

Three new papers ingested by strc-lit-watch (2026-04-27 + 2026-04-28 cron runs). None move S/A directly:

  • 2026-04-24 Xu et al., eLife, “The dual molecular identity of vestibular kinocilia bridges structural and functional traits of primary and motile cilia” (PMID 42029006, DOI 10.7554/eLife.108071, PMC13108966). Auto-tagged delivery 10/10; manual read says vestibular HCs, not cochlear OHCs — score is inflated. Useful background on motile-cilia gene expression in HC bundles for h12 sonogenetic / general bundle context. Not S/A relevant; recommend manual relabel delivery 10/10 → context 4/10.
  • 2026-04-27 Bouzaïd et al., Int J Audiol, “Transient disappearance of OAEs after conventional hearing aid use in OTOF-related auditory neuropathy” (PMID 42037272, DOI 10.1080/14992027.2026.2650503). Auto-tagged applicable 8/10. Useful as h01 paper context for “OAE as OHC-function readout” discussion; not parameter-bearing for any S/A model. Tangential.
  • 2026-03-25 Caragli & Martini, Audiology Research, “Embryonic and Postnatal Gene Therapy Approaches for GJB2-Related Deafness: A Scoping Review” (PMID 42041962, DOI 10.3390/audiolres16020049). Auto-tagged delivery 4/10; review on GJB2 — not STRC-relevant. Score correct; archive without ingest beyond stub.

Pipeline path inconsistency (see Sanity check): all three ingested into ~/STRC/papers/ not ~/STRC/sources/lit/. AGENTS schema (post-2026-04-25) says sources/lit/ is canonical; lit-watch hasn’t been migrated.

Open lit gaps

Carried forward from prior audits, no new gaps closed or opened this run:

  • h09: WH2 × F-actin side-binding Kd — confirmed not measured in primary literature (Chereau 2005 Fig. 1C ITC values only cover G-actin head-binding). Remaining hard blocker; scenario value in models. Hub lit_audit: field = partial.
  • h05: 6 load-bearing parameters still without primary lit (STRC mRNA t½, STRC protein t½, STRC copy-number per OHC, etc.). Paused.
  • h02: piezo parameters (V_saturation, OHC membrane resistance, 210 kPa Young’s modulus, TM displacement 5–30 nm) — partial sourcing; no compute advance until closed.
  • h12 / h13 / h14 / h15 / h16: deferred audit status held.

Egor decisions needed

  1. h01 deliv-tier reconciliation. Hub excerpt records “deliv 3→4” from Phase 8h-lite RWM evidence (~60% applied conc basal turn at 90 min, 7.6× TMPA permeability). Hub deliv: frontmatter still reads 3. Per AGENTS §0 “no silent re-tiering” — please confirm: (a) promote deliv: 3 → 4 and write the missing ## Ranking delta into Phase 8h-lite proof + prepend log entry, or (b) excerpt is overstated, edit it back to deliv: 3 semantics. Synthesis-only run refused to choose.
  2. h01 wetlab RFQ green light. Three RFQ packets staged in wetlab_handoff_2026-04-27/: (i) custom synthesis of 1-indanyl_acylsulfonamide_SO2Me_-Cl + adamantyl_acylsulfonamide_SO2Me_-Cl (+ optional adamantyl_-CF3 residence comparator); (ii) STRC 1066-1216 K1141-fragment WT/E1659A expression feasibility; (iii) DSF/nanoDSF + SPR/BLI/MST direct-binding cascade. Send all three? Or sequence (i) → (ii) → (iii) to control burn?
  3. h03 Ultra-Mini gBlock + B8+WPRE3 cassette order. S-tier next step has been “place order” since 2026-04-24 audit closed all blockers. Standing question; no new compute will move this — order placement is the gate.
  4. h09 reviewer audience. Reviewer pack drafted (premortem + precedent table + orthogonal WH2 controls + transfer memo). Decide: (a) external academic reviewer (Holt lab? specific name?), (b) internal critical pass first, (c) hold and run Phase 4l for paper-grade C_eff before external review.
  5. Lit-watch path migration. ~/bin/strc-lit-watch.py writes to ~/STRC/papers/; AGENTS canonical post-2026-04-25 is ~/STRC/sources/lit/ with type: lit + kind: schema. Three new files (Xu, Bouzaïd, Caragli) sit in the deprecated path. Approve a one-time migration + writer-path patch, or leave both paths active?
  6. AF3 Phase 7H carryover. 12-job batch at ~/STRC/models/af3_jobs_2026-04-24_phase7h/ listed as “blocked on Google auth” in 2026-04-25 morning report. Submitted? Killed? If still pending, ~5 min of browser submission unlocks orthogonal cross-check on 6 NSAID-class parents.
  7. autoresearch/phase5e_v2_rho_optimizer. Scaffold present (program.md, eval.py, baseline.log, history.jsonl); not registered in scripts inventory under that path. Ratify the directory and add the inventory rows, or roll it back?

Compute summary

None — synthesis only. No GPU/MPS/CPU work this run.

Sanity check

  • Wikilink balance, this file: [[ count = ]] count (verified pre-write).
  • Hub frontmatter ↔ ## Ranking delta audit: h09 PASS (Phase 4m promotion at 2026-04-27 15:22 has paired Ranking delta + log prepend + hub frontmatter tier: tentative-S → S); h01 FAIL — excerpt claims deliv 3→4 from Phase 8h-lite without a paired Ranking delta or frontmatter edit. Surfaced as Egor decision #1, not auto-fixed.
  • Scripts inventory rows for new phase scripts (Phase 5q, 8h-lite, 9b geometry / min-relax, 4n MC, 4m soft-floor, 4p Boltz-2, etc.): NOT VERIFIED this run — full inventory is 94 KB; spot-check deferred. Recommend brain-lint pass + visual scan next run. The autoresearch/phase5e_v2_rho_optimizer scaffold is definitively not registered (see decision #7).
  • Lit-watch destination path: writes to ~/STRC/papers/, not canonical ~/STRC/sources/lit/. See decision #5.
  • Routine carry-forward gap: 2026-04-26 and 2026-04-27 had significant proof shipments (h01 Phase 5q, h09 Phase 4m promotion, wetlab handoff) that were never wrapped into a morning report. Today’s report folds them in retroactively; future runs should fire daily so this gap doesn’t reopen.

Connections

Graph View

Backlinks