Add static DDI predictors (fm, kobs, AUCR reversible/TDI, induction) and PK/TK helper columns

[bgunnepalli]

#Use cases
-Many users currently export PKNCA results and compute DDI predictors in Excel or ad-hoc scripts; adding them natively improves customizability and reduces errors.
-These metrics are widely used in IND/NDA DDI screening packages and toxicology reports.

Proposed parameters (initial set)

1. fm — Fraction metabolized (per enzyme) — CLint_enzyme / sum_CLint — depends: CLint_enzyme, CLint_total
2. kobs — TDI observed inactivation rate — kinact * I / (KI + I) — depends: kinact, KI, I (I from I_Cmax_u or I_inlet)
3. aucr_reversible — Predicted reversible inhibition AUCR — 1 / ((1 - fm) + fm/(1 + I/Ki)) — depends: fm, I, Ki
4. aucr_tdi — Predicted AUCR for TDI (mechanistic approx) — 1 / ((1-fm) + fm*(kdeg/(kdeg + kobs))) — depends: fm, kobs, kdeg
5. induction_factor — Induction fold-change — 1 + (Emax * I)/(EC50 + I) — depends: Emax, EC50, I
6. aucr_net — Combined net effect (inhibition/induction) — aucr_reversible / induction_factor (or composite model) — depends: above
7. I_Cmax_u — Unbound peak perpetrator concentration — fu * Cmax — depends: fu, Cmax

Units & molecular-weight (MW) considerations

DDI calculations require consistent concentration units. Users may have PK data in mass units (ng/mL) while in vitro Ki/IC50/EC50 are often reported in molar units (µM, nM).

[billdenney]

The suggested parameters here are mostly related to in vitro work. For that, the invitroTKstats package may be a better fit (https://cran.r-project.org/package=invitroTKstats).

In my experience, typical DDI parameters for noncompartmental analysis are AUC and Cmax ratios for the substrate drug (which do not require molecular weight conversion) and metabolic ratios for metabolites (which require molecular weight conversion).

We do want to add unbound calculations to support organ impairment studies (e.g. hepatic and renal impairment). For that, we would need to support an additional source column for fraction unbound.