Skip to main content

One Codex API client and Python library

Project description

One Codex API - Python Client Library and CLI

Circle CI codecov Black Code Style

Command line interface (CLI) and Python client library for interacting with the One Codex v1 API (API docs).

MAINTAINERS: @polyatail, @bovee, @boydgreenfield

Installation

This package provides 3 major pieces of functionality: (1) a core Python client library; (2) a simple CLI for interacting with the One Codex platform that uses that core library; and (3) optional extensions to the client library, which offers many features aimed at advanced users and provides functionality for use in interactive notebook environments (e.g., IPython notebooks).

Basic installation

The CLI (and core Python library) may be simply installed using pip. To download a minimal installation (#1 and #2), simply run:

pip install onecodex

Installation with optional extensions

To also download the optional extensions to the client library - and all of their dependencies - use the command pip install onecodex[all]. Warning: Because other packages used in the extensions rely upon numpy being present during their installation, numpy must be installed seperately first. So if you do not have numpy installed, and you are going to install onecodex[all] please do the following:

# If numpy is not installed in your environment
pip install numpy

# Once you have numpy installed
pip install onecodex[all]

Using the CLI

Logging in

The CLI supports authentication using either your One Codex API key or your One Codex username and password. To log in using your username and password:

onecodex login

This command will save a credentials file at ~/.onecodex, which will then automatically be used for authentication the next time the CLI or Python client library are used (OS X/Linux only). You can clear this file and remove your API key from your machine with onecodex logout.

In a shared environment, we recommend directly using your One Codex API key, rather than logging in and storing it in a credentials file. To use API key authentication, simply pass your key as an argument to the onecodex command:

onecodex --api-key=YOUR_API_KEY samples

Your API key can be found on the One Codex settings page and should be 32 character string. You may also generate a new API key on the settings page in the web application. Note: Because your API key provides access to all of the samples and metadata in your account, you should immediately reset your key on the website if it is ever accidentally revealed or saved (e.g., checked into a GitHub repository).

Uploading files

The CLI supports uploading FASTA or FASTQ files (optionally gzip compressed) via the upload command.

onecodex upload bacterial_reads_file.fq.gz

Multiple files can be uploaded in a single command as well:

onecodex upload file1.fq.gz file2.fq.gz ...

You can also upload files using the Python client library:

uploaded_sample1 = ocx.Samples.upload("/path/to/file.fastq")

# Or upload a tuple of paired end files
uploaded_sample2 = ocx.Samples.upload(("/path/to/R1.fastq", "/path/to/R2.fastq"))

Which returns a Samples resource (as of 0.5.0). Samples can be associated with tags, metadata, and projects at upload timing using those respective keyword arguments:

# Note format must match the schema defined for our API, with arbitrary
# metadata allowed as a single-level dictionary in the `custom` field.
# See https://docs.onecodex.com/reference#the-metadata-resource for details.
metadata = {
    "platform": "Illumina NovaSeq 6000",
    "date_collected": "2019-04-14T00:51:54.832048+00:00",
    "external_sample_id": "my-lims-ID-or-similar",
    "custom": {
        "my-string-field": "A most interesting sample...",
        "my-boolean-field": True,
        "my-number-field-1": 1,
        "my-number-field-2": 2.0,
    }
}

Uploads can be made in parallel using Python threads (or multiple processes), e.g.:

import concurrent.futures
uploaded_samples = []

with concurrent.futures.ThreadPoolExecutor(max_workers=4) as executor:
    futures = {executor.submit(ocx.Samples.upload, file) for file in LIST_OF_FILES}
    for future in concurrent.futures.as_completed(futures):
        try:
            uploaded_samples.append(future.result())
        except Exception as e:
            print("An execption occurred during your upload: {}".format(e))

Resources

The CLI supports retrieving your One Codex samples and analyses. The following resources may be queried:

  • Your samples (Samples)

  • Sample metadata (Metadata)

  • Analyses, which include several subtypes with additional functionality and fields:

    • Classifications, which are basic metagenomic classification results for your samples
    • Panels, which are in silico panels for particular genes or other functional markers (example on One Codex)
  • Jobs, which provide information on the name, version, and type of analysis which was performed for a given Analyses

Simply invoke the onecodex command, using one of the above resource names as a subcommand (all lowercase). For example:

# fetch all your samples
onecodex samples

# fetch a list of panels based on their ids
onecodex panels 0123456789abcdef 0987654321fdecba

Using the Python client library

Initalization

To load the API, use the following import:

from onecodex.api import Api

Instantiate an API client either by passing your API key or automatically fetching your credentials from ~/.onecodex if you've previously called onecodex login.

from onecodex.api import Api

# Instantiate a One Codex API object, will attempt to get credentials from ~/.onecodex
ocx = Api()

# Instantiate an API object, manually specifying an API key
ocx = Api(api_key="YOUR_API_KEY_HERE")

Resources

Resources are exposed as attributes on the API object. You can fetch a resource directly by its ID or you can fetch it using the query interface. Currently you can access resources using either get() or where(). If you need help finding the ID for a sample, its identifier is part of its url on our webpage: e.g. for an analysis at https://app.onecodex.com/analysis/public/1d9491c5c31345b6, the ID is 1d9491c5c31345b6. IDs are all short unique identifiers, consisting of 16 hexadecimal characters (0-9a-f).

sample_analysis = ocx.Classifications.get("1d9491c5c31345b6")   # Fetch an individual classification
sample_analysis.results()  # Returns classification results as JSON object
sample_analysis.table()    # Returns a pandas dataframe

In addition to methods on individual instances of a given resource (e.g., a Sample or an Analysis), the library also provides methods for aggregating sets of samples or analyses:

all_completed_analyses = ocx.Classifications.where(complete=True)
all_completed_analyses.to_otu()   # Returns classification results as JSON object
all_completed_analyses.to_df()    # Returns a pandas dataframe

Development

Before developing, git and python (version >=2.7 and/or >=3.4) are needed. We recommend using pyenv for Python version management.

To download the client library from GitHub:

git clone https://github.com/onecodex/onecodex.git
cd onecodex/

To set up the project, first create a virtual environment and then install dependencies:

virtualenv venv
source venv/bin/activate
pip install numpy  # numpy must be installed before any of its dependencies
pip install -r requirements.txt

Test are run through the Makefile, and call tox. Note this may take awhile at first because of installing dependencies:

make lint
make test

We use pre-commit for automated linting using black, flake8, and various whitespace and newline formatters during development.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for onecodex, version 0.7.2
Filename, size File type Python version Upload date Hashes
Filename, size onecodex-0.7.2.tar.gz (127.5 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page