Combined CP0 Systems Model — Dietary Rosmarinic Acid + Engineered DAF SCR1-4 (Computational, comp-029)¶
Frozen analysis lives at
./etc/experiments/comp-029-combined-cp0-systems-model/— README + analyze.py + inputs/ + outputs/ all committed for reproducibility. This wiki page is the interpretive layer; the analysis script is stdlib-only Python 3 and reproduces deterministically with RNG seed 29.
The question¶
complement-c5a-gout.md §9.7 added the combined-CP0-coverage hypothesis on 2026-05-15: two CP0 threads (dietary rosmarinic acid + engineered DAF SCR1-4) operate at different geometric scales (fluid-phase / gut-luminal vs. MSU crystal surface), use different mechanisms (covalent C3b modification vs. decay-acceleration), and might therefore compose into additive CP0 coverage without relying on avacopan (the only OE CP0 candidate with serious clinical-development infrastructure, but expensive and on-patent).
Does the predicted combined-effect range, with explicit confidence bounds over published priors, meaningfully exceed either intervention alone?
Verdict: YELLOW¶
| DAF accessibility prior | DAF α | RA alone median | DAF alone median | Combined median | Combined/better-singleton ratio | Verdict |
|---|---|---|---|---|---|---|
| Low | 0.05 | 0.886 | 0.568 | 0.956 | 1.08 | YELLOW |
| Medium | 0.20 | 0.886 | 0.815 | 0.979 | 1.10 | YELLOW |
| High | 0.80 | 0.886 | 0.914 | 0.989 | 1.08 | YELLOW |
Per the comp-029 brief decision rule (GREEN needs combined ≥ 1.5× the better singleton AND combined 95% CI strictly above singleton 95% CI), none of the three accessibility priors clears either threshold. The combined median is only 1.08-1.10× the better singleton, and the combined 95% CI overlaps both singleton CIs.
Why YELLOW, mechanistically¶
Both arms individually saturate at their operative concentrations. RA in the gut-luminal regime (50-1100 μM after 50-200 mg oral dose) hits the C3 convertase IC50 range (5-180 μM) hard — median inhibition fraction 0.886. DAF SCR1-4 at 10-500 nM secreted concentration with even modest accessibility hits decay fraction 0.5-0.9 across the three accessibility priors. When two interventions each already produce 60-90% reduction independently, multiplicative composition (the standard treatment of two independent fractional reductions: f_combined = 1 - (1-f_RA)(1-f_DAF)) has structurally little room to grow — the combined ratio against the better singleton is mathematically capped near 1.1×.
This is not a refutation of the combined-coverage hypothesis. It's a mathematical feature of saturated singletons. The two interventions DO compose as the §9.7 hypothesis claimed; the issue is that under current input uncertainty, the combined effect is probably larger than either alone but not provably-large-enough to justify wet-lab co-administration cost.
What comp-029 rules OUT¶
The RED verdict path is closed. No interaction blocker found at any of three candidate failure modes:
- RA covalently inactivates DAF SCR1-4 via the catechol / α,β-unsaturated ester — no published evidence. RA's covalent chemistry targets activated C3b nucleophilic acceptor sites; SCR/CCP-domain proteins do not present the same chemistry, and DAF binds C3b through reversible protein-protein interaction.
- DAF degraded under gut-luminal RA-bioavailable conditions — comp-012 explicitly verified DAF SCR1-4 LOW protease risk in shio-koji and gut-luminal conditions. RA does not alter pH or protease activity in a way that would compromise DAF stability.
- RA + DAF compete for C3b binding — RA modifies C3b post-deposition (covalent); DAF accelerates C2a/Bb dissociation from already-assembled C3b complexes. The two operate at different time points on different C3b states. Mechanistically complementary, not competitive.
(Mechanistic Extrapolation.)
What comp-029 rules IN¶
-
The two interventions hit the geometric scales the §9.7 hypothesis claimed. RA contributes via the gut-luminal transient, not systemic plasma — Baba 2004 free plasma Cmax of 20 nM is ~1700× below the central IC50 (34 μM), giving essentially 0% inhibition (median 0.0007). The gut-luminal regime at 252-1100 μM (Kang 2021) DOES reach mechanistically active concentrations — median 0.886. This sharpens the wet-lab framing: the right RA readout is a gut-luminal complement-activation assay, not a plasma-based one. (Mechanistic Extrapolation over In Vitro Sahu 1999 + Clinical Trial Baba 2004 anchors.)
-
DAF SCR1-4 contributes via MSU-surface decay-acceleration, with effective concentration in the 10-500 nM range and accessibility α as the single load-bearing wet-lab unknown (comp-012's explicit Limitations section).
-
The combined-effect distribution is broadly higher than either singleton, just not separated under current prior uncertainty. The wide singleton CIs (RA p5=0.44, DAF p5=0.18-0.77 depending on α) leak into the combined CI.
Dominant uncertainty drivers¶
| Input | Spearman r vs combined (α=0.20) | Prior spread |
|---|---|---|
| RA gut-luminal IC50 | -0.658 (largest) | 36× (5-180 μM) |
| RA gut-luminal [RA] | +0.552 | 22× (50-1100 μM) |
| DAF effective [DAF] | +0.480 | 50× (10-500 nM); accessibility α adds another 16× |
Top wet-lab measurement that would tighten the prediction most:
Measure DAF SCR1-4 MSU-surface engagement fraction. This is the load-bearing wet-lab unknown per comp-012's explicit Limitations section and the dominant uncertainty driver in comp-029's sensitivity analysis. One experiment, one parameter, currently a 16× prior spread.
validation-experiments.md §1.25 DAF SCR1-4 expression screen already provides the right functional readout (zymosan or MSU-crystal C5a-generation assay ± DAF). The §1.25 functional arm directly resolves α as a side product of its primary experimental aim. No new wet-lab experiment is gated by comp-029 — only an optional co-treatment arm WITHIN the already-planned §1.25 experiment, gated on §1.25's own functional readout.
Wet-lab handoff to §1.25¶
If §1.25 returns α ≥ 0.5 (mid-to-high MSU-surface engagement), comp-029 re-runs to GREEN — DAF alone saturates and the combined ratio improves. The marginal-cost RA co-treatment arm is then justified:
Optional co-treatment arm (gated on §1.25 functional readout showing DAF SCR1-4 mid-to-high MSU-surface engagement): Add a co-treatment condition to the §1.25 zymosan / MSU-crystal C5a-generation assay: DAF SCR1-4 expressed material (mature, ELISA-quantified) + rosmarinic acid at 100 μM (gut-luminal regime; matches the median Sahu 1999 C3b-deposition IC50) co-incubated 30 min at 37°C with complement-competent serum + MSU crystals. Compare C5a + C5b-9 generation across (i) DAF SCR1-4 alone, (ii) RA alone, (iii) DAF + RA combined, (iv) vehicle. Marginal cost: one additional condition per plate; readout: same ELISA already in the §1.25 plan.
If §1.25 returns α < 0.2, comp-029 stays YELLOW and the combined-strategy thesis is parked. RA alone via the gut-luminal regime survives as a stand-alone dietary intervention; engineered DAF SCR1-4 survives as a stand-alone CP0 candidate at the MSU surface; the combined-coverage claim is shelved until a new mechanism or geometry surfaces.
Pass 3 softening discipline honored¶
This page explicitly does not claim:
- The two interventions are "mechanistically additive." Multiplicative composition of independent fractional reductions is the standard treatment; super-additivity would require coupling we did NOT find.
- Rosmarinic acid "saturates fluid-phase and gut-luminal C3 convertase." It saturates the GUT-LUMINAL regime (post-meal, transient), and contributes essentially 0% at SYSTEMIC free-RA concentrations. The two are 4-5 orders of magnitude apart.
- Downstream clinical effect size. Even if the mechanism prediction holds, downstream clinical SUA / IL-1β effect-sizes are gated by the H08-class clinical-translation question covered separately.
Evidence levels¶
- In silico systems composition over published kinetic constants: Mechanistic Extrapolation.
- Sahu 1999 / Englberger 1988 IC50 values: In Vitro.
- Baba 2004 human plasma Cmax: Clinical Trial (single-dose human PK).
- Wang 2017 rat oral absolute bioavailability: Animal Model.
- Kang 2021 gut-luminal concentration calculation: Mechanistic Extrapolation (calculation from oral dose + intestinal volume assumptions).
- Fischer 1981 C4b2a half-life: In Vitro.
- Medof 1984 DAF effective threshold: In Vitro.
Cross-references¶
- comp-018 upstream complement modulator sweep (the dietary thread origin)
- comp-020 upstream complement verification re-run (where the 44× assay-format IC50 spread is documented)
- comp-012 DAF SCR1-4 truncated construct (the engineered surface-decay thread origin; the wet-lab MSU-surface engagement unknown)
- comp-030 DAF cassette ranking (cassette-design refinement)
- chaperone-orthogonal-stacking.md §3.5.3 (DAF effective PDI load)
- H05 DAF SCR1-4 CP0 thesis (the formal hypothesis card)
validation-experiments.md§1.25 (the gating wet-lab experiment)- complement-c5a-gout.md §9.7 (originating section, 2026-05-15)