Published PK and PK/PD models used in the analyses
A catalog of the published population PK and PK/PD models reimplemented, fitted or reused in the analyses on this site, with source papers, structure, what was done and the article.
Models from the literature that the analyses on this site reimplemented, fitted to public data or reused with published estimates. All analyses use public information or simulation only and are research and learning analyses, not dosing proposals. Whether a model was reimplemented from a paper or estimated from public data is stated for each row, because the two support different claims (see reimplementing a published population PK model).
| Drug and model | Source | Structure | What the analysis did | Article |
|---|---|---|---|---|
| Mosunetuzumab population PK (intravenous) | Bender et al., 2024; estimates also checked in Li et al., 2025 | Two-compartment, time-varying clearance, CD20 binding | Reimplemented in Python; step-up regimen exposure compared; RO to CRS link examined as exploratory | Reimplementing a published mosunetuzumab model |
| Pembrolizumab population PK | Ahamadi et al., 2017 (Table 3) | Two-compartment, linear, time-invariant clearance, body-weight exponents | Reimplemented in R (rxode2); exposure by weight band for flat, weight-based and 400 mg Q6W regimens | Fixed doses of anti-PD-1 antibodies · explorer |
| Nivolumab population PK | Bajaj et al., 2017 (Table 1) | Two-compartment, linear, time-varying clearance, body-weight exponents | Reimplemented in R (rxode2); exposure by weight band for flat, weight-based and 480 mg Q4W regimens | Fixed doses of anti-PD-1 antibodies · explorer |
| Tarlatamab population PK | Published fixed effects from the FDA multidisciplinary review (2024); Kong et al., 2025 | Linear two-compartment | Used the published estimates to compare regimens at equal average exposure | Can matching average exposure alone set the tarlatamab dosing interval? |
| Tarlatamab, from published PK summary values | Public summary PK values; Kong et al., 2025 for context | Linear two-compartment (4 parameters fitted to 6 summary values) | Fitted to summary values, not reimplemented; examined what the values determine and what depends on assumptions | How far can public summary PK data evaluate a change in the dosing interval? |
| Nimotuzumab population PK | Data nimoData (nlmixr2data); context in Rodríguez-Vera et al., 2015 |
Three candidates: linear one-compartment, linear two-compartment, two-compartment with nonlinear elimination | Estimated from public concentration data with nlmixr2; asked which structure is needed to compare exposure across dosing conditions; not a reproduction of the original model | How complex does a population PK model need to be? |
| Warfarin PK/PD | Data warfarin (nlmixr2data); Wright et al., 2011 for the PK/PD approach |
Oral one-compartment with lag; turnover-type response | Estimated from public data; asked when individual information improves prediction | When does individual information improve PK/PD prediction? |
| RMC-6236 (RAS inhibitor) | Jiang et al., 2024, Supplementary Table S8 | Oral one-compartment blood PK, tumor exchange, indirect DUSP6 response | Point estimates reused from the supplement; target-coverage scenarios computed | RAS inhibitor dosing: from public PK/PD data to target-coverage scenarios |
Notes
- Agreement with published values shows that a model was rebuilt correctly. It does not validate the model for predicting another population or regimen.
- Where a source contained an inconsistency (for example a table value that differs from its footnote), the articles report it and test both versions.
- The distribution file of the anti-PD-1 analysis contains the code, results and an independent recalculation.
Related: identifiability of PK/PD models and the R package pkident.