iscc.treatment.Treatment ¶

Treatment(adaptive=False, start=0, duration=None, dosage_decay=0.5, rounds=4, rate_multiplier=2.0, toxicity=0.1, effectiveness=0.9, kill_rate=1.5, max_tumor_size=100000, sites='both')

Bases: object

Shared base for the therapies; defines the dosing schedule and per-cell effect.

The therapies (Chemotherapy, TargetedTherapy, Immunotherapy, Surgery) share the dosing / scheduling parameters below; each subclass adds its own targeting knobs and defines how a dose modifies a cell. Pass an instance to a tumour engine's grow(..., treatment=...).

Parameters:

Name	Type	Description	Default
`adaptive`	`bool`	If `True`, dose only while the tumour exceeds `max_tumor_size` (adaptive therapy); if `False` (default), dose continuously within the active window.	`False`
`start`	`int`	First step at which the therapy is active (default 0).	`0`
`duration`	`int`	Number of steps the therapy stays active (default `None` = until the run ends).	`None`
`dosage_decay`	`float`	Between-round dose decay factor (default 0.5).	`0.5`
`rounds`	`int`	Number of dosing rounds (default 4).	`4`
`rate_multiplier`	`float`	Factor applied to a fully-sensitive cell's affected rate under full dose (default 2.0); the exponent depends on the cell's `treatment_resistance`.	`2.0`
`toxicity`	`float`	Per-step probability of off-target action on a non-cancer cell (default 0.1).	`0.1`
`effectiveness`	`float`	Per-step probability the therapy acts on a targeted cell (default 0.9).	`0.9`
`kill_rate`	`float`	Death hazard imposed on a fully-sensitive cell under full dose in the genotype engine (default 1.5); set above the `max_birth_rate` cap so even high-fitness (driver-amplified) sensitive clones still regress while resistant clones escape.	`1.5`
`max_tumor_size`	`int`	Size threshold that gates dosing when `adaptive=True` (default 100000).	`100000`
`sites`	`(both, primary, met)`	Compartment(s) the therapy acts on (default `"both"` = systemic): `"primary"` (neoadjuvant / local) or `"met"` (adjuvant after primary resection).	`"both"`

Methods:

Name	Description
`discrete_event`	One-shot state edits that are NOT rate modifiers (e.g. surgical resection). The base

discrete_event ¶

discrete_event(tumor, step)

One-shot state edits that are NOT rate modifiers (e.g. surgical resection). The base treatment does nothing; Surgery overrides this. The engine calls it once per step, so chemo-only / no-surgery runs stay byte-identical (no state change, no rng draw).