encodeproject 1.0.17

Python package wrapping some of the encode project APIs.

Python package wrapping some of the encode project APIs.

There is a short Notebook with a tutorial available here.

How do I install this package?

As usual, just download it using pip:

pip install encodeproject

Tests Coverage

Since some software handling coverages sometimes get slightly different results, here’s three of them:

Usage Examples

The package contains both methods to run queries on the Encode Project APIs and methods to filter the responses. Every available method has a comprehensive docstring attached to it, so I welcome you to read the source code.

Queries

The library currently offers to query methods that already integrate some filtering properties: one for the experiments and one for the biosamples.

For querying the experiments you can run the following:

from encodeproject import experiment

experiments = experiment()

Let’s take a look to an in-depth example, showing all the available parameters:

from encodeproject import experiment

experiments = experiment(
    # The cell line we are interested in.
    # For example values could be K562 or GM12878.
    # We use None to specify that we are not
    # interested in any particular cell line.
    cell_line = None,
    # The reference genomic assembly we want.
    # For example values could be hg19 or GRCh38
    # We use None to specify that we are not
    # interested in any particular genomic assembly.
    assembly = None,
    # The target (the genes coding for proteins in this context) we want.
    # For example values could be CTCF or H3K27ac
    # We use None to specify that we are not
    # interested in any particular target.
    target = None,
    # The status of the data we want.
    # We only want released data, meaning data that are
    # neither old (archived) or with errors (revoked).
    status = 'released',
    # The organism we are considering.
    # Since we only want Homo sapiens data,
    # we specify that organism name.
    organism = 'Homo sapiens',
    # The format of the files we are interested in
    file_type = 'bigWig',
    # We ask to consider only experiments with replicas
    replicated = True,
    # We only want with the signals
    # expressed as "fold change over control"
    searchTerm = "fold change over control",
    # We do not need to specify any other specific
    # additional parameters
    parameters = None,
    # We want to download all the
    # available experiments
    limit = 'all',
    # We want to drop all the experiments
    # which have been characterized by significand issues
    drop_errors = (
        'extremely low read depth',
        'missing control alignments',
        'control extremely low read depth',
        'extremely low spot score',
        'extremely low coverage',
        'extremely low read length',
        'inconsistent control',
        'inconsistent read count'
    )
)

All parameters are optional, they just act as additional filters.

For querying the biosamples you can run the following:

from encodeproject import biosample

my_biosample_query_response = biosample(
    accession="ENCSR000EDP", # The accession code for the desired biosample
)

As for the experiments there are a number of filters available:

hg19_samples = biosamples(
    # The list of accessions to retrieve
    accessions=accession_codes,
    # Wethever to convert the results in dataframe.
    # The following filters only apply if dataframes are used
    to_dataframe = True,
    # The status of the data we want.
    # We only want released data, meaning data that are
    # neither old (archived) or with errors (revoked).
    status = "released",
    # The organism we want.
    organism = "human",
    # The genomic assembly we want to use
    assembly = "hg19",
    # The output type we want.
    output_type = "fold change over control",
    # And finally the bare minimum amount
    # of biological replicates
    min_biological_replicates = 2
)

For running multiple queries for biosamples at once you can run the following:

from encodeproject import biosamples

responses = biosamples(
    accessions=["ENCSR000EDP", "ENCSR030EDP", "ENCSR067EDP"], # The accessions code for the desired biosamples
)

Filters

Since the response files can get quite big and hard to read, I’ve prepared also a couple filter functions.

For filtering the accessions codes from an experiment response you can use:

from encodeproject import accessions

codes = accessions(my_experiment_query_response)

For filtering the download URLs from a biosample response you can use:

from encodeproject import download_urls

codes = download_urls(my_biosample_query_response)

Utilities

Download utility

I’ve added also a method to download from a given URL, showing a loading bar, based on this answer from StackOverflow.

from encodeproject import download

download("https://encode-public.s3.amazonaws.com/2012/07/01/074e1b37-2be1-4f6a-aa42-6c512fd1834b/ENCFF000XOW.bigWig")

Sample to DataFrame instruction

Utility to convert a sample to a relatively simple pandas DataFrame.

from encodeproject import biosample_to_dataframe

df = biosample_to_dataframe(my_biosample_query_response)

Issues and Feature Requests

This library started out of necessity to script some queries on the encodeproject. If you need some specific feature that isn’t currently already offered by the library, please do proceed with a pull request (quickest way: add the feature yourself and push it on the library) or alternatively you can open an issue and when I’ll get the time I’ll see to it.