HuBMAPy

A package to interact with the HuBMAP Human Reference Atlas Ontology. The package can be used to perform ontology queries programmatically or through a command-line interface.

Installation

Clone the repository:

git clone https://github.com/ccb-hms/HuBMAPy.git

Install package using pip on the cloned repository folder:

pip3 install .

Command Line Interface

hubmapy -q QUERY [-o OUTPUT] [-n NAME]

To display a help message with descriptions of tool arguments do:

hubmapy -h or hubmapy --help

Required arguments

-q QUERY Input query file containing a single SPARQL query.

Optional arguments

-o OUTPUT Path to output folder for the query results files.

-n NAME Name of the query, to be included in the generated query results file.

Programmatic Usage

Import the HuBMAPy library:

import hubmapy

Then use as follows:

hubmap = hubmapy.HuBMAPy()

From here various built-in queries can be performed, for example:

df = hubmap.biomarkers_for_all_cell_types()

where df is a data frame containing the query results.

Additionally, it is possible to perform arbitrary queries specified by users:

df = hubmap.do_query_from_file(query_file_path='...')

where query_file_path is an absolute path to a SPARQL query file. Or:

df = hubmap.do_query(query='...', query_name='...')

where query is a string containing a SPARQL query.

Built-in Queries

The package supports the following built-in queries:

• biomarkers_for_all_cell_types()
• biomarkers_for_all_cell_types_in_anatomical_structure(anatomical_structure=obo:UBERON_0002371) # bone marrow
• biomarkers_for_cell_type_in_anatomical_structure(cell_type=obo:CL_0000787, anatomical_structure=obo:UBERON_0002371) # memory B cell, bone marrow
• tissue_blocks_in_anatomical_structure(anatomical_structure=obo:UBERON_0002048)
• tissue_block_count_for_all_anatomical_structures()
• anatomical_structures_in_tissue_block(tissue_block='...')
• locations_of_all_cell_types()
• evidence_for_specific_cell_type(cell_type=obo:CL_0002394) # i.e., publication DOIs
• evidence_for_all_cell_types()
• cell_types_from_biomarkers(biomarkers=[hgnc:633,hgnc:800])

Dependencies

• ROBOT v1.8.3 (https://github.com/ontodev/robot)
• CCF Ontology v2.0.0-alpha.18 (https://github.com/hubmapconsortium/ccf-ontology)