epigenomic-dataset 1.1.7

Python package wrapping ENCODE epigenomic data for a number of reference cell lines.

Python package wrapping ENCODE epigenomic data for several reference cell lines.

How do I install this package?

As usual, just download it using pip:

pip install epigenomic_dataset

Tests Coverage

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

Preprocessed data for cis-regulatory regions

We have already downloaded and obtained the max window value for each promoter and enhancer region for the cell lines A549, GM12878, H1, HEK293, HepG2, K562 and MCF7 in the dataset Fantom and cell lines A549, GM12878, H1, HepG2 and K562 for the Roadmap dataset taking in consideration all the target features listed in the complete table of epigenomes.

The thresholds used for classifying the activations of enhancers and promoters in Fantom are the default explained in the sister pipeline CRR labels which handles the download and preprocessing of the data from Fantom and Roadmap.

Dataset	Cell line	Promoters	Enhancers
Fantom	A549	200	1000	200	1000
Fantom	GM12878	200	1000	200	1000
Fantom	H1	200	1000	200	1000
Fantom	HEK293	200	1000	200	1000
Fantom	HepG2	200	1000	200	1000
Fantom	K562	200	1000	200	1000
Fantom	MCF-7	200	1000	200	1000
Roadmap	A549	200	1000	200	1000
Roadmap	GM12878	200	1000	200	1000
Roadmap	H1	200	1000	200	1000
Roadmap	HepG2	200	1000	200	1000
Roadmap	K562	200	1000	200	1000

Here are the labels for all the considered cell lines.

Dataset	Promoters	Enhancers
Fantom	200	1000	200	1000
Roadmap	200	1000	200	1000

TODO: align promoters and enhancers in a reference labels dataset.

The complete pipeline used to retrieve the CRR epigenomic data is available here.

Automatic retrieval of preprocessed data

You can automatically retrieve the data as follows:

from epigenomic_dataset import load_epigenomes

X, y = load_epigenomes(
    cell_line = "K562",
    dataset = "fantom",
    regions = "promoters",
    window_size = 200,
    root = "datasets" # Path where to download data
)

Pipeline for epigenomic data

The considered raw data are from this query from the ENCODE project

You can find the complete table of the available epigenomes here. These datasets were selected to have (at time of the writing, 07/02/2020) the least possible amount of known problems, such as low read resolution.

You can run the pipeline as follows: suppose you want to extract the epigenomic features for the cell lines HepG2 and H1:

from epigenomic_dataset import build

build(
    bed_path="path/to/my/bed/file.bed",
    cell_lines=["HepG2", "H1"]
)

If you want to specify where to store the files use:

from epigenomic_dataset import build

build(
    bed_path="path/to/my/bed/file.bed",
    cell_lines=["HepG2", "H1"],
    path="path/to/my/target"
)

By default, the downloaded bigWig files are not deleted. You can choose to delete the files as follows:

from epigenomic_dataset import build

build(
    bed_path="path/to/my/bed/file.bed",
    cell_lines=["HepG2", "H1"],
    path="path/to/my/target",
    clear_download=True
)