BED classifier tutorial

BED classifier is a utility that allows you to classify BED files based on the number of columns and the types of data contained within those columns.

Information on various data formats can be found here: https://genome.ucsc.edu/FAQ/FAQformat.html

Additional detailed specifications for the Browser Extensible Data (BED) format can be found here: https://samtools.github.io/hts-specs/BEDv1.pdf

get_bed_classification

The function, get_bed_classification, takes a path to bed-like file or a dataframe and returns a BedClassification object with the following attributes:

class BedClassification(BaseModel):
    bed_compliance: str
    data_format: DATA_FORMAT
    compliant_columns: int
    non_compliant_columns: int

where DATA_FORMAT is defined as:

class DATA_FORMAT(str, Enum):
    UNKNOWN = "unknown_data_format"
    UCSC_BED = "ucsc_bed"
    BED_RS = "bed_rs"
    BED_LIKE = "bed_like"
    BED_LIKE_RS = "bed_like_rs"
    ENCODE_NARROWPEAK = "encode_narrowpeak"
    ENCODE_NARROWPEAK_RS = "encode_narrowpeak_rs"
    ENCODE_BROADPEAK = "encode_broadpeak"
    ENCODE_BROADPEAK_RS = "encode_broadpeak_rs"
    ENCODE_GAPPEDPEAK = "encode_gappedpeak"
    ENCODE_GAPPEDPEAK_RS = "encode_gappedpeak_rs"
    ENCODE_RNA_ELEMENTS = "encode_rna_elements"
    ENCODE_RNA_ELEMENTS_RS = "encode_rna_elements_rs"

Example usage of the BED classifier:

from bedboss.bedclassifier.bedclassifier import get_bed_classification

classification = get_bed_classification("path/to/bedfile.bed")


print(f"{classification.bed_compliance}, {classification.data_format}, {classification.compliant_columns}, {classification.non_compliant_columns}")


## Example 1
## > 'bed3+0', 'ucsc_bed', 3, 0

## Example 2
## > 'bed6+4', 'encode_narrowpeak', 6, 4

Data formats

Below rs refers to relaxed_score which indicates that a fifth column was present where the values are integers greater than 0. In constrast, a strict interpretation for column 5 is:

Column 5 - score - A score between 0 and 1000.

UNKNOWN

Classification was unable to determine the data format.

UCSC_BED

Conforms to ucsc bed

BED_RS

Conforms to ucsc bed but with a relaxed interpretation for the fifth column.

BED_LIKE

Data is tab delimited but contains columns that are not compliant with ucsc bed. Example: bedn+m where n are compliant columns, m are non-compliant columns and m > 0

BED_LIKE_RS

Data is tab delimited but contains columns that are not compliant with ucsc bed but with a relaxed interpretation for the fifth column.

Example: bedn+m where n are compliant columns, m are non-compliant columns and m > 0, Column 5 = integer > 0

ENCODE_NARROWPEAK

Conforms to ENCODE narrowPeak

ENCODE_NARROWPEAK_RS

Conforms to ENCODE narrowPeak but with a relaxed interpretation for the fifth column.

ENCODE_BROADPEAK

Conforms to ENCODE broadPeak

ENCODE_BROADPEAK_RS

Conforms to ENCODE broadPeak but with a relaxed interpretation for the fifth column.

ENCODE_GAPPEDPEAK

Conforms to ENCODE gappedPeak

ENCODE_GAPPEDPEAK_RS

Conforms to ENCODE gappedPeak but with a relaxed interpretation for the fifth column.

ENCODE_RNA_ELEMENTS

Conforms to ENCODE RNA elements

ENCODE_RNA_ELEMENTS_RS

Conforms to ENCODE RNA elements but with a relaxed interpretation for the fifth column.