profile · writing · software · linkedin
I'm Daiki Kumakura, a mathematical biologist and bioinformatics researcher with a Ph.D. in Life Science from Hokkaido University.
I work with dynamic models and quantitative data analysis. My current interests include PK/PD, pharmacometrics, and what data can — and cannot — tell us about a model.
R / Python / ODEs / statistical modeling / reproducible analysis
| Project | What it does |
|---|---|
| pkident | Structural and practical identifiability for PK/PD ODE models. R package, in development. |
| CRiSM | Concatenated ribosomal marker sequences for phylogenetic analysis. Python prototype. |
| RLR_transform | Research scripts for compositional time series and Convergent Cross Mapping. |
| ShotgunMetagenomics | Container-based scripts for read processing and functional profiling. |
Research articles and reproducible analyses on public data, model estimation, diagnostics, uncertainty, and the limits of a conclusion. For example:
- When does individual information improve PK/PD prediction? — warfarin PK/PD reanalysis
- Reimplementing a published mosunetuzumab population PK model — step-up dosing exposure (in Japanese)
- Can matching average exposure alone set the tarlatamab dosing interval? — population PK simulation (in Japanese)
- Do early tumor measurements improve prediction of subsequent survival? — landmark analysis
All analyses by topic: population PK, PK/PD, dose selection, tumor dynamics and immune biomarkers.