BioGrade · expert reasoning data for biopharma agentsEnvironmentsBenchmarkCompare modelsData flywheelThesis
← First-in-Human Starting-Dose Selection comparison · model: GPT (frontier)
LLM-judge score: 5.0 / 5 · strong
allometric_scaling: 5 species_selection: 5 safety_mabel_judgment: 5 final_dose: 5 faithfulness: 5
The agent correctly applied FDA BSA scaling with the right km factors (rat 6, dog 20, human 37), computed both HEDs accurately (1.62 and 2.70 mg/kg), selected rat as the governing most-sensitive species, applied the 10x safety factor for MRSD (0.162 mg/kg), and reached ~9.7 mg total at 60 kg — matching the reference answer. For a non-agonist small molecule, NOAEL/MRSD governs and MABEL is not required, which the agent handled appropriately. All numbers trace faithfully to the tool data with no fabrication.
Clinical Pharmacology / Translational · Clinical pharmacologist

First-in-Human Starting-Dose Selection — agent trajectory for expert review

gpt-5.2-chat-latest · 2026-06-06T10:29:25.344521+00:00 · 3 tool calls · 7.2s · 2197 tokens
run id: fih_dose_20260606_031816_fih-dose__gpt-frontier

Query

Recommend a first-in-human starting dose for ZX-101, an oral small-molecule MEK1/2 inhibitor. Use the GLP tox NOAELs and FDA body-surface-area scaling; show your HED per species, the governing species, the MRSD, and the total mg dose at 60 kg.

Trajectory

Agent · turn 1
⚙ calls get_molecule(molecule=ZX-101)
⚙ calls get_tox_noael(molecule=ZX-101)
⚙ calls get_km_factors()
↳ get_molecule result
show data
{
  "id": "ZX-101",
  "name": "ZX-101",
  "modality": "Small molecule (oral kinase inhibitor)",
  "target": "MEK1/2",
  "high_risk": false,
  "high_risk_reason": null,
  "human_weight_kg": 60.0,
  "notes": "Direct cytotoxic/anti-proliferative MoA; tox-driven dose selection is appropriate. No immune-agonist risk, so NOAEL->HED->MRSD with a 10x safety factor governs.",
  "most_appropriate_species_note": "Standard FDA practice: convert each species, then take the lowest HED (most sensitive) unless a species is known to be a poor model."
}
↳ get_tox_noael result
show data
{
  "id": "ZX-101",
  "name": "ZX-101",
  "units": "mg/kg/day (GLP repeat-dose NOAEL)",
  "species_noael_mg_kg": {
    "rat": 10.0,
    "dog": 5.0
  },
  "species_not_studied": [
    "monkey"
  ],
  "human_weight_kg": 60.0,
  "most_appropriate_species_note": "Standard FDA practice: convert each species, then take the lowest HED (most sensitive) unless a species is known to be a poor model."
}
↳ get_km_factors result
show data
{
  "description": "FDA 2005 body-surface-area (km) conversion factors. HED (mg/kg) = animal NOAEL (mg/kg) x (animal km / human km). Human adult km = 37.",
  "human_km": 37,
  "default_safety_factor": 10,
  "default_human_weight_kg": 60.0,
  "km_factors": {
    "human_adult": {
      "km": 37,
      "ref_weight_kg": 60.0
    },
    "human_child": {
      "km": 25,
      "ref_weight_kg": 20.0
    },
    "mouse": {
      "km": 3,
      "ref_weight_kg": 0.02
    },
    "hamster": {
      "km": 5,
      "ref_weight_kg": 0.08
    },
    "rat": {
      "km": 6,
      "ref_weight_kg": 0.15
    },
    "ferret": {
      "km": 7,
      "ref_weight_kg": 0.3
    },
    "guinea_pig": {
      "km": 8,
      "ref_weight_kg": 0.4
    },
    "rabbit": {
      "km": 12,
      "ref_weight_kg": 1.8
    },
    "dog": {
      "km": 20,
      "ref_weight_kg": 10.0
    },
    "monkey": {
      "km": 12,
      "ref_weight_kg": 3.0
    },
    "marmoset": {
      "km": 6,
      "ref_weight_kg": 0.35
    },
    "squirrel_monkey": {
      "km": 12,
      "ref_weight_kg": 0.6
    },
    "baboon": {
      "km": 20,
      "ref_weight_kg": 12.0
    },
    "micro_pig": {
      "km": 27,
      "ref_weight_kg": 20.0
    },
    "mini_pig": {
      "km": 35,
      "ref_weight_kg": 40.0
    }
  }
}
Agent · turn 2

Step 1 — GLP Repeat‑Dose NOAELs

Human Km = 37 (60‑kg adult)
HED formula (FDA 2005):

[
\textbf{HED (mg/kg)} = \text{Animal NOAEL (mg/kg)} \times \left(\frac{K_m\text{ animal}}{37}\right)
]

Step 2 — Convert to Human Equivalent Dose (HED)

Rat

Km (rat) = 6

[
10 \times \frac{6}{37} = 1.62 \text{ mg/kg}
]

Dog

Km (dog) = 20

[
5 \times \frac{20}{37} = 2.70 \text{ mg/kg}
]

Step 3 — Governing Species

Lowest HED = 1.62 mg/kg (rat)

→ Rat is the governing species

Step 4 — Apply FDA Default Safety Factor (10×)

[
\textbf{MRSD (mg/kg)} = \frac{1.62}{10} = 0.162 \text{ mg/kg}
]

Step 5 — Convert to Total Daily Dose (60‑kg Human)

[
0.162 \times 60 = 9.7 \text{ mg}
]

✅ Recommended First‑in‑Human Starting Dose

Parameter Value
Governing species Rat
HED (rat) 1.62 mg/kg
MRSD 0.162 mg/kg
Total daily dose (60 kg) ~10 mg

📌 Final Recommendation

Proposed FIH starting dose: 10 mg orally once daily

This is:
- Based on the most sensitive species (rat)
- Derived via FDA BSA scaling
- Includes the standard 10× safety factor
- Appropriate for a small‑molecule oncology MEK inhibitor where NOAEL‑based MRSD is standard

If you'd like, I can also project exposure margins versus expected efficacious exposure to sanity‑check whether this is likely to be pharmacologically active or strictly sub‑therapeutic.

Final output

Step 1 — GLP Repeat‑Dose NOAELs

Human Km = 37 (60‑kg adult)
HED formula (FDA 2005):

[
\textbf{HED (mg/kg)} = \text{Animal NOAEL (mg/kg)} \times \left(\frac{K_m\text{ animal}}{37}\right)
]

Step 2 — Convert to Human Equivalent Dose (HED)

Rat

Km (rat) = 6

[
10 \times \frac{6}{37} = 1.62 \text{ mg/kg}
]

Dog

Km (dog) = 20

[
5 \times \frac{20}{37} = 2.70 \text{ mg/kg}
]

Step 3 — Governing Species

Lowest HED = 1.62 mg/kg (rat)

→ Rat is the governing species

Step 4 — Apply FDA Default Safety Factor (10×)

[
\textbf{MRSD (mg/kg)} = \frac{1.62}{10} = 0.162 \text{ mg/kg}
]

Step 5 — Convert to Total Daily Dose (60‑kg Human)

[
0.162 \times 60 = 9.7 \text{ mg}
]

✅ Recommended First‑in‑Human Starting Dose

Parameter Value
Governing species Rat
HED (rat) 1.62 mg/kg
MRSD 0.162 mg/kg
Total daily dose (60 kg) ~10 mg

📌 Final Recommendation

Proposed FIH starting dose: 10 mg orally once daily

This is:
- Based on the most sensitive species (rat)
- Derived via FDA BSA scaling
- Includes the standard 10× safety factor
- Appropriate for a small‑molecule oncology MEK inhibitor where NOAEL‑based MRSD is standard

If you'd like, I can also project exposure margins versus expected efficacious exposure to sanity‑check whether this is likely to be pharmacologically active or strictly sub‑therapeutic.

Reference answer key (deterministic) — open after grading

Deterministic FDA-MRSD / MABEL calc over the curated nonclinical inputs. HED = NOAEL x (animal km / human km); most sensitive species governs; MRSD = HED / 10; total = MRSD x 60 kg. For high-risk agonists, MABEL is computed from in vitro potency / receptor occupancy and the LOWER dose is recommended (the TGN1412 lesson). A strong agent should reproduce these steps and units.

StepValueNote
MoleculeZX-101 (Small molecule (oral kinase inhibitor))target MEK1/2
HED from rat10.0 x 6/37 = 1.622 mg/kgNOAEL 10.0 mg/kg, km 6
HED from dog5.0 x 20/37 = 2.703 mg/kgNOAEL 5.0 mg/kg, km 20
Governing species (most sensitive)rat -> HED 1.622 mg/kglowest HED across studied species
MRSD (NOAEL path)1.622 / 10 = 0.162 mg/kgdefault 10x safety factor
NOAEL-based total dose0.162 mg/kg x 60.0 kg = 9.73 mg60.0 kg human
RECOMMENDED FIH starting dose9.73 mg (via NOAEL/HED)governed by NOAEL/HED

Expert grading

HED / allometric scaling correctness 1 (poor) 5 (excellent)
5: Applies HED = NOAEL x (animal km / human km) with the correct km factors (rat 6, dog 20, monkey 12, human 37) and computes each species' HED accurately.
1: Uses wrong km values, inverts the ratio, scales by simple mg/kg without BSA correction, or makes arithmetic errors in the HED.
Most-sensitive-species selection 1 (poor) 5 (excellent)
5: Selects the governing species correctly — the lowest HED (most sensitive) for the small molecule, and the pharmacologically relevant species (NHP) for the antibodies — and justifies it.
1: Picks the highest HED, ignores a more sensitive species, or uses a non-binding rodent species for a human-specific antibody.
Safety factor & MABEL judgment 1 (poor) 5 (excellent)
5: Applies the default 10x safety factor, and for immunostimulatory/agonist biologics computes a MABEL from potency/receptor occupancy and takes the LOWER of MABEL vs NOAEL — explicitly invoking the TGN1412 rationale.
1: Omits the safety factor, dosing a high-risk agonist off the NOAEL alone, or fails to compute/compare MABEL when it clearly governs.
Final dose reasonableness & units 1 (poor) 5 (excellent)
5: States a single recommended starting dose with correct units (mg or ug total at 60 kg), in a plausible range — mg-range for the small molecule/mAb, microgram-range for the high-risk biologics.
1: Reports a dose with wrong/missing units, off by orders of magnitude, or recommends a dangerously high dose for an agonist (e.g. a TGN1412-style 0.1 mg/kg).
Evidence faithfulness 1 (poor) 5 (excellent)
5: Every number traces to the returned NOAELs, km table, and pharmacology; no fabricated NOAELs, potencies, or species.
1: Invents NOAEL or potency values, uses km factors not in the table, or contradicts the returned tool data.
Overall verdict
Your name / ID