Bioframe: Operations on Genomic Interval Dataframes

Bioframe is a library to enable flexible and scalable operations on genomic interval dataframes in python. Building bioframe directly on top of pandas enables immediate access to a rich set of dataframe operations. Working in python enables rapid visualization (e.g. matplotlib, seaborn) and iteration of genomic analyses.

The philosophy underlying bioframe is to enable flexible operations: instead of creating a function for every possible use-case, we instead encourage users to compose functions to achieve their goals.

Bioframe implements a variety of genomic interval operations directly on dataframes. Bioframe also includes functions for loading diverse genomic data formats, and performing operations on special classes of genomic intervals, including chromosome arms and fixed size bins.

Read the docs, including the guide.

If you use bioframe in your work, please cite via its zenodo DOI 10.5281/zenodo.5703622

Installation

The following are required before installing bioframe:

• Python 3.7+
• numpy
• pandas>=1.3

pip install bioframe

Interval operations

Key genomic interval operations in bioframe include:

• closest: For every interval in a dataframe, find the closest intervals in a second dataframe.
• cluster: Group overlapping intervals in a dataframe into clusters.
• complement: Find genomic intervals that are not covered by any interval from a dataframe.
• overlap: Find pairs of overlapping genomic intervals between two dataframes.

Bioframe additionally has functions that are frequently used for genomic interval operations and can be expressed as combinations of these core operations and dataframe operations, including: coverage, expand, merge, select, and subtract.

To overlap two dataframes, call:

import bioframe as bf

bf.overlap(df1, df2)

For these two input dataframes, with intervals all on the same chromosome:

overlap will return the following interval pairs as overlaps:

To merge all overlapping intervals in a dataframe, call:

import bioframe as bf

bf.merge(df1)

For this input dataframe, with intervals all on the same chromosome:

merge will return a new dataframe with these merged intervals:

See the guide for visualizations of other interval operations in bioframe.

File I/O

Bioframe includes utilities for reading genomic file formats into dataframes and vice versa. One handy function is read_table which mirrors pandas’s read_csv/read_table but provides a schema argument to populate column names for common tabular file formats.

jaspar_url = 'http://expdata.cmmt.ubc.ca/JASPAR/downloads/UCSC_tracks/2018/hg38/tsv/MA0139.1.tsv.gz'
ctcf_motif_calls = bioframe.read_table(jaspar_url, schema='jaspar', skiprows=1)

Tutorials

See this jupyter notebook for an example of how to assign TF motifs to ChIP-seq peaks using bioframe.

Projects currently using bioframe:

• cooler
• cooltools
• yours? :)