SimpleColumnMapper

ColumnMapper for columns that contain information about a single phenotypic abnormality only. This kind of ColumnMapper should be used for columns that can be representing by one HPO term and which can contain symbols such as "+", "Y", or "yes" indicating that the abnormality was observed, symbols such as "-", "N", or "no" indicating that the abnormality was explicity excluded, and (optionally) symbols indicating that the abnormality was not measured or assessed.

For instance, the following mapper would generate a phenotypic feature for Global developmental delay if the column contains "yes" (observed) and would call it as excluded if the column contains "no". For any other text or for an empty cell, the feature would be called as not measured (and would not be included in the phenopacket).

SimpleColumnMapper constructor

from pyphetools.creation import SimpleColumnMapper

ddMapper = SimpleColumnMapper(hpo_id='HP:0001263',
    hpo_label='Global developmental delay',
    observed='yes',
    excluded='no')

It can be convenient to add multiple SimpleColumnMappers at the same time. The following function enables this. Note that the HpoParser object from the pyphetools.creation package is used to create a concept recognizer object. The column_d dictionary is used to store the individual mappers, and will be passed later on to a CohortMapper object.

Creating multiple SimpleColumnMapper objects at once

import hpotk

onto_store = hpotk.configure_ontology_store()
hpo = onto_store.load_hpo()

from pyphetools.creation import HpoExactConceptRecognizer

hpo_cr = HpoExactConceptRecognizer.from_hpo(hpo)

##
column_mapper_d = {}
items = {
    'regression': ["Developmental regression", "HP:0002376"],
    'autism': ['Autism', 'HP:0000717'],
    'hypotonia': ['Hypotonia', 'HP:0001252'],
    'movement disorder': ['Abnormality of movement', 'HP:0100022'],
    'CVI': ['Cerebral visual impairment', 'HP:0100704'],
    'seizures': ['Seizure', 'HP:0001250'],
    'DD': ['Global developmental delay', 'HP:0001263']
}
item_column_mapper_d = hpo_cr.initialize_simple_column_maps(column_name_to_hpo_label_map=items,
                                                            observed='yes',
                                                            excluded='no')
# Transfer to column_mapper_d
for k, v in item_column_mapper_d.items():
    column_mapper_d[k] = v