Finding set relationships between two or more genomic datasets is an important problem. As an example, one might want to know how much overlap exists between two sets of signal peak calls, to quantitatively establish the degree to which they are similar.
The
bedops program offers several set and multiset operations, including union, subset, and difference, to assist investigators with answering these types of questions.
This program takes sorted UCSC BED-formatted data as input, either from a file or streamed from standard input. If your data are unsorted, use
sort-bed or
bbms to prepare data for consumption by
bedops. You only need to sort once!
The
bedops program is highly scalable. Because memory usage is constant, one can use sorted inputs of open-ended size with an according linear increase in processing time.
Visit the Google Code site for usage information and a description of set operations and other run-time options:
http://code.google.com/p/bedops/wiki/bedops
