nmdc-client 1.0.0

Python client library for the NMDC API (formerly nmdc-api-utilities)

nmdc-client

Renamed: This package was previously published on PyPI as nmdc-api-utilities. The old name still installs (and pulls in nmdc-client automatically with a DeprecationWarning), but new development happens here. Please migrate your installs to pip install nmdc-client and your imports to from nmdc_client import ... at your earliest convenience.

A Python client for the NMDC (National Microbiome Data Collaborative) Runtime API. NMDC is a multi-institutional effort to enable findable, accessible, interoperable, and reusable (FAIR) microbiome data. This library provides Pythonic access to NMDC metadata so you can query, filter, and traverse linked records without writing raw API calls. For more information on the NMDC Runtime API itself, see the API documentation.

What you can do

• Search and retrieve metadata for studies, biosamples, data objects, workflow executions, and more
• Traverse linked records across collections (e.g. study → biosamples → data objects)
• Apply MongoDB-style filters to build precise queries
• Work with geospatial filters for applicable records (e.g. biosample collection locations)
• Access privileged submission and staging endpoints (authorized users only)

Documentation

Full documentation is at microbiomedata.github.io/nmdc-client.

For additional real-world examples, the nmdc_notebooks repository contains Jupyter notebooks that use this package end-to-end.

Installation

pip install nmdc-client

To stay up to date with new releases:

pip install --upgrade nmdc-client

Quick start

Single record lookup

from nmdc_client import BiosampleSearch

biosample_client = BiosampleSearch()
biosample_client.get_record_by_id(record_id="nmdc:bsm-13-amrnys72")

Multi-step workflow: study → biosamples → data objects

A more realistic workflow starts from a study name and traverses linked records:

from nmdc_client import StudySearch, DataObjectSearch

# Step 1: Find a study by name
study_client = StudySearch()
studies = study_client.get_record_by_attribute(
    attribute_name="name",
    attribute_value="Molecular mechanisms underlying changes in the temperature sensitive "
                    "respiration response of forest soils to long-term experimental warming",
    exact_match=True,
)

study_id = studies[0]["id"]  # e.g. "nmdc:sty-11-8ws97026"

# Step 2: Retrieve all biosamples linked to that study
biosamples = study_client.get_linked_instances(
    ids=[study_id],
    types=["nmdc:Biosample"],
    hydrate=True,
)
biosample_ids = [b["id"] for b in biosamples]

# Step 3: Retrieve data objects linked to those biosamples
data_object_client = DataObjectSearch()
data_objects = data_object_client.get_linked_instances(
    ids=biosample_ids,
    types=["nmdc:DataObject"],
    hydrate=True,
)

get_linked_instances is available on all search clients and accepts any list of NMDC IDs and target record types.

Filtering with MongoDB-style operators

from nmdc_client import BiosampleSearch

client = BiosampleSearch()

# Find biosamples from soil
results = client.get_record_by_filter(
    filter={
        "env_medium.has_raw_value": {"$regex": "soil"},
    }
)

See the MongoDB filters guide for supported operators and more examples.

Available search clients

All clients share a common interface (get_record_by_id, get_record_by_attribute, get_record_by_filter, get_records, get_linked_instances).

Biosample & study
for retrieving metadata about biosamples, studies, and collection sites
BiosampleSearch, StudySearch, CollectingBiosamplesFromSiteSearch, FieldResearchSiteSearch

Sample processing
for retrieving metadata about sample processing and storage
ProcessedSampleSearch, MaterialProcessingSearch, StorageProcessSearch

Data generation
for retrieving metadata about data generation processes, instruments, and configurations for sequencing and mass spectrometry
DataGenerationSearch, ManifestSearch, InstrumentSearch, ConfigurationSearch

Data & workflow
for retrieving metadata about workflows for processing raw data, data objects, and functional annotations
WorkflowExecutionSearch, DataObjectSearch, CalibrationSearch, FunctionalAnnotationAggSearch

See the CollectionSearch subclasses reference for full detail on each class.

Public vs. privileged API

Most users will only need the public search clients above. A separate privileged API is available for users with NMDC submission and staging permissions. See the Privileged API reference for details.

Migrating from nmdc-api-utilities

If you were previously installing nmdc-api-utilities, both of the following will continue to work:

• pip install nmdc-api-utilities — installs the tombstone shim, which pulls in nmdc-client and emits a DeprecationWarning.
• from nmdc_api_utilities import ... and from nmdc_api_utilities.<submodule> import ... — still resolve, with a DeprecationWarning.

nmdc-api-utilities 0.7.0 is the final release of that distribution name — we won't be republishing it. The shim is frozen at the API surface that existed at the time of the rename: existing submodules will keep tracking the latest nmdc-client 1.x (so bug fixes and improvements to those submodules reach you transparently), but any new submodule added to future nmdc-client releases will only be accessible under the new name. To pick up new functionality, you'll need to migrate.

To migrate fully:

pip uninstall nmdc-api-utilities
pip install nmdc-client

Then replace nmdc_api_utilities with nmdc_client in your imports.

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines on reporting issues, suggesting features, and submitting pull requests.

License

This project is licensed under the terms in LICENSE.