Inputs¶

Overview¶

The main input abstraction is GeneConfig, either provided directly or constructed from a pandas dataframe using CancerRLTOModel.from_dataframe().

A secondary input is Microenvironment, which encodes environmental stress conditions.

Gene-level inputs¶

Required field¶

Only one dataframe column is strictly required beyond identifiers:

gene
transcription_rate

Optional fields¶

The code supports the following optional columns:

Column	Type	Default	Meaning
`translation_efficiency`	float	`1.0`	Translation gain factor
`mrna_half_life`	float	`6.0`	mRNA half-life
`protein_half_life`	float	`12.0`	Protein half-life
`copy_number`	float	`2.0`	Effective gene dosage
`clone_fraction`	float	`1.0`	Fraction of clones carrying the gene state
`essentiality_weight`	float	`1.0`	Weight on robustness benefit
`threshold_abundance`	float or null	inferred	Explicit critical abundance threshold
`baseline_abundance`	float or null	inferred	Explicit baseline abundance override
`burden_weight`	float	`0.02`	Linear burden weight
`toxicity_weight`	float	`0.0`	Superlinear toxicity weight
`regulation_strength`	float	`0.0`	Stabilizing regulation term
`burstiness`	float	`1.0`	Noise control used in gamma parameterization
`stress_sensitivity`	float	`0.2`	Environmental stress sensitivity
`oncogenic_boost`	float	`0.0`	Abundance boost from oncogenic pressure
`pathway`	str	`generic`	Pathway/grouping label

Microenvironment inputs¶

Microenvironment defines five bounded stress axes:

hypoxia
nutrient_limitation
immune_pressure
drug_pressure
oxidative_stress

Each value is clipped to [0, 1].

Validation rules¶

GeneConfig.validate() enforces:

transcription_rate > 0
translation_efficiency > 0
copy_number > 0
clone_fraction in [0, 1]
essentiality_weight >= 0
burden_weight >= 0
toxicity_weight >= 0
burstiness > 0

Invalid values raise ValueError.

Example input table¶

gene,transcription_rate,essentiality_weight,burden_weight,toxicity_weight,pathway
MYC,8.0,0.8,0.015,0.020,proliferation
KRAS,5.5,0.9,0.012,0.015,MAPK
PASSENGER,2.0,0.0,0.01,0.01,none

Demo dataset¶

The module ships with CancerRLTOModel.demo_dataset(), a synthetic stress-test dataset containing:

typical anchor genes,
zero-essentiality and zero-penalty edge cases,
noise extremes,
burden and toxicity extremes,
microenvironment-sensitive genes.

Warning

If your real project expects omics-derived inputs, sample metadata, or file-based ingestion beyond pandas dataframe construction, that behavior is not present in the supplied module and should be documented only after implementation.