pyhealth.tasks.CancerMutationBurden

class pyhealth.tasks.CancerMutationBurden(code_mapping=None)

Bases: BaseTask

Task for predicting high vs low tumor mutation burden.

This task classifies patients based on their tumor mutation burden (TMB), which is associated with immunotherapy response. TMB is approximated by counting the number of mutated genes.

task_name

The name of the task.

Type:

str

input_schema

The input schema specifying required inputs.

Type:

Dict[str, str]

output_schema

The output schema specifying outputs.

Type:

Dict[str, str]

TMB_THRESHOLD

Mutation count threshold for high TMB classification.

Type:

int

Note

This is a simplified TMB calculation based on gene count. Clinical TMB is typically measured as mutations per megabase of sequenced DNA.

Examples

>>> from pyhealth.datasets import TCGAPRADDataset
>>> from pyhealth.tasks import CancerMutationBurden
>>> dataset = TCGAPRADDataset(root="/path/to/tcga_prad")
>>> task = CancerMutationBurden()
>>> samples = dataset.set_task(task)

task_name: str = 'CancerMutationBurden'

input_schema: Dict[str, str] = {'age_at_diagnosis': 'tensor', 'mutations': 'sequence'}

output_schema: Dict[str, str] = {'high_tmb': 'binary'}

TMB_THRESHOLD: int = 10

pre_filter(df)

Return type:

LazyFrame