Add modular QSP plugin system with turnover biomarker model

[jam-sudo]

Motivation

• Add a modular QSP layer so mechanistic biomarker/cytokine models can be run driven by PBPK concentration signals.
• Support both a simple post-hoc workflow and an optional coupled PBPK+QSP solve to enable tighter integration when needed.

Description

• Introduce a new physio_sim.qsp package with BaseQSPModel interface, a model registry (register_qsp_model, get_qsp_model, list_qsp_models), and a built-in turnover model (physio_sim/qsp/models/turnover.py).
• Add QSP schema and loader (QSPConfig, load_qsp) in physio_sim/config.py and an example YAML examples/qsp_turnover.yaml describing kin, kout, emax, ec50_mg_per_L, hill, and optional B0.
• Extend the solver (physio_sim/pbpk/solver.py) to support posthoc (default) QSP solving using interpolated plasma signal and an optional coupled mode that augments PBPK state vector and solves PBPK+QSP together with solve_ivp(method="BDF") (kept optional behind qsp_mode).
• Add CLI non-breaking flags in physio_sim/cli.py: --qsp-model, --qsp-config, and --qsp-mode posthoc|coupled, and append QSP state columns (e.g., B_biomarker) to timecourse.csv and add Bmax, t_Bmax, Bend to summary.json when QSP is enabled.
• Add documentation (docs/qsp.md, docs/qsp_models.md), README quickstart snippet, and unit/cli tests (tests/test_qsp_turnover.py, tests/test_cli_qsp_smoke.py).

Testing

• Ran formatting and linting: python -m ruff format . and python -m ruff check . which completed without failures.
• Ran static typing checks: python -m mypy physio_sim which completed with no issues reported.
• Ran the test suite: python -m pytest -q and all tests passed (18 passed).

---

Codex Task

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: d05a97a079

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[jam-sudo]

I applied PR #7 conceptually on by manually merging its intended changes (QSP plugin CLI integration + turnover model coupling path). Resolved conflicts with current branch state, validated locally (, , on touched files), then closed PR #7 since upstream content is now represented in commit 0382ac9 on main.

[jam-sudo]

I applied PR #7 intent on main (QSP plugin CLI integration + turnover model coupling path) by manual merge due conflict. Conflicts were resolved against current branch state, validated locally (pytest, mypy, ruff on touched files), and committed as 0382ac9 on main. Closing PR #7 as superseded by manual merge.