Paircomp - Comparing Two Genomes to Find Conserved cis-regulatory Segments

Paircomp - Comparing Two Genomes to Find Conserved cis-regulatory Segments


With sequencing prices falling rapidly, it has now become possible to sequence the genomes and transcriptomes of multiple related species. This allows biologists to compare genome sequences of different related species, and find regions conserved over large evolutionary distance. We came across a really cool software tool for comparing genomes (paircomp) developed by Eric Davidson’s group in Caltech.

This program takes two sequences (supposedly regions around the same gene from two species) and generates plots like these -

In the above figure, two axes show genomic coordinates from two organism. The program takes a k-mer value as input (let’s say 21), and compares all 21-mers from one sequence with all 21-mers from the other. If the 21-mers match, the program places a red dot in the graph at the corresponding location. ‘Match’ here can be perfect match, or the user can allow to have some mismatch between the k-mers being compared.

You can also present the same information as the above dot-plot in genomic view:

Two sequences being compared are shown at the top and bottom. A thin line connecting them matching 21-mers between two sequence. When many such lines fall next to each other, you see a thick line representing a conserved patch.

You can download the latest version of paircomp from here. Installation is straightforward, provided you have the latest version of python installed. Older versions are available here along with papers, manuals and the viewer program.

Paircomp is a combination of three modules -

i) A C++ program (paircomp) that compares two sequences and generates a table with many lines (each representing a red dot), each having four columns - coordinate in sequence 1, coordinate in sequence 2, number of matching nucleotides and strand. The program is run as ‘paircomp seq1 seq2 K-mer-size threshold output-file’. Threshold=.8 means any match over 80% similarity will be reported.

ii) A simple python program that converts the above paircomp output into a XML file. This python program adds the sequences and gene annotations if any to the paircomp output.

iii) A viewer program (FamilyRelations) to view the above XML output. The images we showed above are from the viewer program.



Written by M. //