GROUNDING — h12 sonogenetic (proportionate pass, C-tier)
G1 — “Wu et al. 2023”: a bare author-year with no locatable paper
Claim being audited, verbatim from
hypotheses/h12-sonogenetic/phases/STRC Sonogenetics Bifurcation Analysis.md:87:
“Even if fold induction is 3x higher than Wu et al. 2023 (200x instead of 62x), protein stays under 200% of target. The saturation cap dominates.”
and line 156:
”### Wu et al. 2023 — used 6xNFAT only”
Exact command run over the vault’s literature SSoT:
ls sources/lit/ | grep -i "wu\|2023"
Output — three Wu papers, none of them this one:
2026-04-21-wu-pvdf-trfe-pedotpss-piezoelectric.md
2026-04-29-wu-research-progress-of-myo6.md
2026-08-14-wu-acbd7-is-essential-for.md
(PVDF-TrFE piezoelectric material — an h02 source; a MYO6 review; an ACBD7 paper. None is a 6×NFAT sonogenetic-induction study.)
Evidential weight: ABSENCE of the source, not of the phenomenon. The citation carries no PMID, no DOI, no journal, no title — only “Wu et al. 2023” and a number, 62×. That is the exact signature of the phantom citations that the 2026-04-23 and 2026-04-25 audits destroyed in h02 (“Masaki 2009”, “Gueta 2006”, “Ren 2011”) and h05 (“Sharma 2018”, “Krey 2015”, “Chao 2011 PMID 21458670”). It is not asserted here that Wu 2023 is a phantom — it may well be a real paper that simply was never filed. It is asserted that it is unverifiable as it stands, and that h12’s one external quantitative anchor is therefore unchecked.
G2 — the parameters that do NOT need a literature source
Recorded because it changes the blocker count and is easy to get wrong.
Kd_CaN and NFAT_threshold read at first glance like unsourced constants
(the note’s line 47 calls them out: “NFAT_threshold (0.05) and Kd_CaN (500 nM)
are both too low”). But the analysis does not rest on point values — it sweeps
them:
- line 91: “Swept Kd_CaN (100-2000 nM) against NFAT_threshold (0.01-0.30) at 70 dB.”
- line 111: “Swept 245 combinations: Kd_CaN (500-5000 nM) x NFAT_threshold (0.05-0.40) x Hill (2-8).”
A conclusion that holds across a 10–50× swept range does not depend on a literature value for the swept parameter, and demanding one would be audit theatre. These are correctly treated as design/sweep parameters, not physical constants, and are not blockers. The sweep is, in fact, the right way to handle an unmeasured parameter — h12 did by construction what h09 had to reach for retroactively.
G3 — the bifurcation result itself
In-house computation, not a literature parameter: the self-dosing window saturates at ~20 dB, giving 123% of target at a 14 dB whisper and 198% at 70 dB (table at lines 39–43). No external source is required for a model’s own output; what would require sourcing is any constant fed into it. Per G2, the sensitive ones were swept, and per G1 the one external anchor is unverified.