intersect

This program is a CLI for chr.Intersect package. The program finds homologous regions common for a set of closely related input nucleotide sequences.

Compile

Given that you have git, golang, and make:

git clone https://github.com/IvanTsers/intersect
cd intersect
make

Tutorial

Change to tutorial:

cd tutorial

There are 10 sequences containing from 1 to 4 marker regions. We run intersect using t1.fasta as a reference and sequences in t as queries:

intersect -r t1.fasta -d t

This will return a perfect intersection, that is, a region shared by all sequences. It contains only one of the markers. We lower the sensitivity threshold to 0.9:

intersect -r t1.fasta -d t -f 0.9

This returns two markers shared by at leat 9 out of 10 genomes. We continue to lower the sensitivity of intersection and add switches to print mutation positions:

intersect -r t1.fasta -d t -f 0.6 -n

This time the intersection includes three markers, two of them are separated with mismatches shown as N. These regions are found in at least 6 out of 10 genomes. We lower the sensitivity threshold again and add the -verbose flag to print some statistics:

intersect -r t1.fasta -d t -f 0.2 -n -verbose

The intersection now contains all four markers.

To reproduce results of a phylonium -p run

use these parameters:

intersect -r <reference> -d <targets> -s -clean-reference -clean-queries

• the default sensitivity threshold -f is 1.0 (perfect intersection)
• -s prints the positions of segregating sites into headers
• -clean-reference removes non-ATGC nucleotides from the reference
• -clean-queries removes non-ATGC nucleotides from the queries

Use -h to see all available options.