crr-labels 1.1.1

FANTOM

To retrieve the FANTOM promoters and enhancers you can proceed as follows:

from crr_labels import fantom

enhancers, promoters = fantom(
    cell_lines=["HelaS3", "GM12878"], # list of cell lines to be considered.
    window_size=200, # window size to use for the various regions.
    genome = "hg19", # considered genome version. Currently supported only "hg19".
    center_enhancers = "peak", # how to center the enhancer window, either around "peak" or the "center" of the region.
    enhancers_threshold = 0, # activation threshold for the enhancers.
    promoters_threshold = 5, # activation threshold for the promoters.
    drop_always_inactive_rows = True, # whether to drop the rows where no activation is detected for every row.
    binarize = True, # whether to return the data binary-encoded, zero for inactive, one for active.
    nrows = None # the number of rows to read, useful when testing pipelines for creating smaller datasets.
)

The library will download and parse the fantom project raw data and return two DataFrames for the required cell lines. Consider reading the method docstring for more in-depth information about the method.

The main steps are the following:

• The raw files are retrieved from the fantom dataset from the link specified in the fantom_data.json file

• The window for the enhancers and promoters are expanded or compressed to the given window size. In particular:

  • The enhancers’ window can either be centered on the region center with the “center” mode or around the “peak” with the “peak” mode.

  • The promoters’ window is upstream in the positive strand from the end of the promoter and downstream on the negative strand from the start of the promoter.

• When multiple experiments are present for a cell line, for instance for “HelaS3”, an average of the activation peaks is executed.

• Optionally (and by default) the rows that are always inactive for the chosen cell lines are dropped. You can specify this behaviour using the parameter “drop_always_inactive_rows”.

Roadmap

To retrieve the Roadmap promoters and enhancers you can proceed as follows:

from crr_labels import roadmap

enhancers, promoters = roadmap(
    cell_lines = ["HelaS3", "GM12878"], # List of cell lines to be considered.
    window_size = 200, # Window size to use for the various regions.
    genome = "hg19", # Considered genome version. Currently supported only "hg19".
    states = 18, # Number of the states of the model to consider. Currently supported only "15" and "18".
    enhancers_labels = ("7_Enh", "9_EnhA1", "10_EnhA2"), # Labels to encode as active enhancers.
    promoters_labels = ("1_TssA",), # Labels to enode as active promoters.
    nrows = None # the number of rows to read, useful when testing pipelines for creating smaller datasets.
)

Consider reading the method docstring for more in-depth information about the method.

FANTOM

The following datasets contain data for the cell lines GM12878, HelaS3, HepG2, K562, A549, H1, H9, JURKAT, MCF7, HEK293, Caco2, HL60 and PC3.

TODO: Render the datasets

Roadmap

The following datasets contain data for the cell lines GM12878, HelaS3, HepG2, K562, A549, H1, H9, DND41, HUES48, HUES6, HUES64 and IMR90.

TODO: The updated processed labels will be added as soon as we decide on the new states to be used.