Human Model · HM(t) at day 115
State-timeSynthetic demo data⟨ I, St, H0:t, Et, Θt, T, Ut ⟩ — belief, not fact; uncertainty shown, not hiddenSteering view
1
Localization — where is this individual biologically now?
Belief state b_t per dimension: value ± sd from within-person noise plus the observation model's semantic distance. Trajectory features (velocity, autocorrelation, recovery time) are state variables.
| Dimension | b_t | dS/dt | autocorr | recovery | observation model |
|---|---|---|---|---|---|
| Therapy response | 41 ± 8.8 | -0.41/d | 0.56 | 11 d | Imaging + tumour markers, interpolated · proxy · last read ≤ 14 d |
| Gastrointestinal function | 84 ± 7.6 | +0.98/d | 0.69 | 1 d | Daily GI check-in · direct · today |
| Treatment toxicity | 60 ± 7.9 | +0.32/d | 0.66 | 5 d | Labs, interpolated between draws · proxy · last draw ≤ 7 d |
| Inflammatory activity | 70 ± 11.9 | +0.49/d | 0.41 | 1 d | CRP/ferritin + immune signature · inferred · weeks |
| Immune / metabolic | 43 ± 11.7 | -0.4/d | 0.15 | 1 d | Metabolic panel + wearable · inferred · weeks |
| Nutritional sufficiency | 27 ± 8.7 | -0.4/d | 0.57 | 3 d | Weight, albumin, food log · proxy · days |
| Adherence | 43 ± 4.2 | -0.16/d | 0.49 | 4 d | Dose + to-do completion · direct · today |
| Function / quality of life | 21 ± 6.7 | -0.71/d | 0.58 | 4 d | PRO items · direct · today |
High autocorrelation and slow recovery are early-warning signatures of a critical transition (Scheffer et al.), independent of the level.
2
Provenance — how did they arrive here?
The path H_0:t: baseline constraints, divergence, perturbations, actions and responses in order.
- day 0baselineBaseline S06: EC-03 phenotype; Poor metabolizer; butyrate 44
- day 11perturbationLow-HRV stretch (-2.7)
- day 48divergenceTrajectory diverged from the population model
- day 48noteTrajectory diverged from prediction
- day 50recommendedGI-04 recommended
- day 51acceptedPatient accepted
- day 59responseNo meaningful change
- day 59recommendedME-01 recommended
- day 60acceptedPatient accepted
- day 63responseDeterioration detected before clinical escalation
- day 72perturbationDietary drift (-2.6)
- day 75perturbationPoor-sleep week (-3)
- day 95perturbationAlcohol on dosing days (-1.8)
Now: between basins, heading towards failing.
3
Mechanism — which theories best explain the trajectory?
Versioned competing theories with supporting (E+), contradicting (E−) and missing-but-discriminative (E?) evidence. The contradiction engine lists what should NOT happen under each theory and whether it did.
T2 v2leading
confidence 33%
Bile-acid diarrhoea on a low-butyrate ecology: reduced SCFA production leaves bile acids unconjugated after fat-rich meals; GI burden tracks fat intake with a one-day lag.
E+
- Butyrate capacity 44/100
- DeepGene: 6 keystone butyrate producers absent
- Fat load → GI +5.4 next day
- Diversity 2.9
E−
- —
E? · would discriminate
- Faecal bile acids
- 7α-hydroxy-4-cholesten-3-one (C4)
- Rechallenge: 3 high-fat days under observation
Forbidden under this theory
✓ holdsGI should not rise on low-fat days (fat effect 5.4 over 39 low-fat days)
v1 baseline: microbiome → v2 dietary attribution
T1 v3competing
confidence 23%
Mucosal barrier loss: drug-induced epithelial injury amplified by a FUT2-dependent glycan deficit; histamine and inflammatory load add on top.
E+
- DeepGene: Klebsiella pneumoniae above range (LPS / zonulin)
- DeepGene: host DNA fragmentation atypical
- DeepGene: tissue recycling 62
- Inflammatory load 70
E−
- No FUT2 variant
- No histamine signal in the log
E? · would discriminate
- Faecal calprotectin (would separate barrier from motility)
- Serum DAO / histamine
- Zonulin
Forbidden under this theory
✓ holdsGI should not improve while histamine intake stays high (GI velocity 1.15/day)
✓ holdsLow-residue meals should reduce GI within a week (not tested)
v1 baseline: genotype + phenotype → v2 day 51: GI-04 response added to E− → v3 dietary attribution from 60-day log
T4 v1competing
confidence 23%
Behavioural spiral: symptomatic days cut adherence and sleep, which raise GI and stress, which cut adherence further; the biology is a follower, not the driver.
E+
- Adherence 43%
- Sleep quality worsening
- Stress 8.7/10
E−
- —
E? · would discriminate
- Dose-taking timestamps vs symptom timestamps
- Sleep-window rechallenge
Forbidden under this theory
✓ holdsGI should not worsen on days with full adherence and good sleep (insufficient paired days)
v1 from tracker trends
T3 v1weakened
confidence 21%
Drug-exposure toxicity: pharmacokinetic over-exposure drives GI and systemic toxicity together; dietary and microbial factors are secondary.
E+
- Poor metabolizer: exposure above modelled range
- Toxicity 60
- GI autocorrelation 0.84 (slow recovery)
E−
- —
E? · would discriminate
- Trough drug level
- Dose-timing rechallenge (evening vs morning)
Forbidden under this theory
✗ violatedGI should not move with diet if exposure is the driver (3 dietary detractors with ≥ 80% confidence)
v1 baseline: pharmacogenomics
4
Forecast — which futures are likely under alternative actions?
30-day rollout of GI burden under the baseline model T and under each safe candidate action (effect ramps over onset). Dashed = prediction, never fact.
No action
GI 84 → 72
Under NP-02
GI 84 → 56
Under SC-12
GI 84 → 61
Under DT-01
GI 84 → 63
8
Intervention as measurement
Each past action as a measurement of this patient's transition dynamics: predicted ΔGI vs observed, prediction error ε_t, and the within-patient evidence patterns → individualized response confidence (observational, kept separate from randomized evidence).
| Action | pred ΔGI | obs ΔGI | ε | temporality | magnitude | dechallenge | rechallenge | cross-modal | confidence |
|---|---|---|---|---|---|---|---|---|---|
| GI-04 day 51 | -13 | 39 | +52 | ✗ | ✓ | no | n/a | ✗ | 18% |
| ME-01 day 60 | -3 | 21 | +24 | ✗ | ✓ | n/a | n/a | ✗ | 18% |
Learning curve · |ε| per cycle:GI-04 52 (prior 52)ME-01 19 (prior 24)improving over cycles
5
Decision — the Restore controller
Safe set computed first, then A* = argmax E[HealthGain + λ₁·InformationGain + λ₂·ConfidenceGain − λ₃·Risk − λ₄·Burden − λ₅·Cost]. λ₁ = 0.6 (theories near-equiprobable), λ₃ = 1.4 (fragile state). Execution stays with the clinician or the informed patient (EU MDR Class I).
| Class | Action | Health | Info | Conf. | Risk | Burden | Cost | Utility |
|---|---|---|---|---|---|---|---|---|
| Escalate | CR-01 Clinician Review 100% of 4 similar improved. | 0.37 | 0.2 | 0.25 | 0.02 | 0.25 | 0.15 | 0.38 |
| Measure | Faecal bile acids Discriminates T2 from its competitors (E?). | 0 | 0.75 | 0.2 | 0.02 | 0.25 | 0.2 | 0.32 |
| Treat | NP-02 Dietary Support Protocol 62% of 61 similar improved. | 0.45 | 0.2 | 0.25 | 0.13 | 0.25 | 0.15 | 0.31 |
| Test | 7α-hydroxy-4-cholesten-3-one (C4) Discriminates T2 from its competitors (E?). | 0 | 0.75 | 0.2 | 0.02 | 0.25 | 0.5 | 0.26 |
| Treat | SC-12 Supportive Compound 66% of 44 similar improved. | 0.35 | 0.2 | 0.25 | 0.08 | 0.25 | 0.4 | 0.23 |
| Observe | Continue passive collection, no perturbation A clean transition can be observed only without a new perturbation. | 0 | 0.15 | 0.05 | 0 | 0.05 | 0 | 0.08 |
| Treat | HY-01 Hydration Protocol 0% of 0 similar improved. | 0.15 | 0.2 | 0.25 | 0.08 | 0.25 | 0.15 | 0.08 |
| Maintain | Hold current regimen and content Protects nothing while the state deteriorates. | 0.05 | 0.05 | 0.05 | 0.4 | 0.05 | 0 | -0.49 |
Excluded from the safe set
- GI-04 Not indicated for the current state.
- ME-01 Not indicated for the current state.
- DT-01 Dose-timing change requires investigator authorization in poor metabolizers (contraindication from I).
- BR-03 Behavioural-only action while toxicity ≥ 55: worst-case trajectory crosses the safety boundary; escalation required first.
- AD-05 Behavioural-only action while toxicity ≥ 55: worst-case trajectory crosses the safety boundary; escalation required first.
Evidence weights · w = Authority × Relevance × Persistence
- Biological identity Poor metabolizer; CYP2D6Germline panel, baseline0.86 (0.95·0.9·1)
- High-authority clinical Toxicity 60, response 41Labs / imaging, ≤ 14 d0.77 (0.9·1·0.85)
- Intervention response 2 within-patient responses recordedΔS after each action, accumulating0.68 (0.75·1·0.9)
- Deep biological state DeepGene retest day 90: damage 83/100, Shannon 1.67, butyrate 316 CPM, 3 opportunists above, 6 keystone absentSynthetic DeepGene metagenomics + host DNA, sample 910887-R (compositional proxy)0.54 (0.7·1·0.775)
- Dynamic state GI 84, sleep 6.9 h, HRV 34Companion + wearable, today0.18 (0.6·1·0.3)
- Exposure context Prepare a plain, low-fat pasta or rice meal and note how it feels afterwards; dosing per protocol; adherence 43%Food log + dose log, event-level0.1 (0.55·0.9·0.2)
LLIFE OS insight · the five questions, in prose
grounded · add an LLM key for proseP-01454 is on day 115 of NBL-101 (EC-03, baseline S06): response unstable, tolerance unstable, GI unstable. Over 14 days response moved -8 and GI +9; discontinuation risk is 92%. The state is between basins, heading towards failing, after diverging from the model on day 48.
Biology sets the gain: Poor metabolizer, Low microbial diversity, Cluster EC-03 (high response / high intolerance), Adherence < 60%; butyrate capacity 44/100, diversity 2.9. Pushing it down: Enterococcus faecium 16.476% (LPS load) (-34), Low butyrate capacity (-28.1), 6 keystone species absent (-21). From the daily log: fat +5.4 GI next day, alcohol +3.9 GI next day, simple carbs +3.8 GI next day.
Past actions: GI-04 on day 51 → worsened (-4%); ME-01 on day 60 → improved (3%).
Next: NP-02 — 77% of 284 similar trajectories improved, 70% probability here. This would be contradicted by a further GI rise within the onset window or falling adherence.