README file for AGE software distribution

ABOUT
Software AGE implements optimal algorithms for aligning genomic sequences with
structural variations (SVs). Precise alignment allows for correct breakpoint
determination. The algorithms are described in the publication
Bioinformatics. 2011 Mar 1;27(5):595-603. Epub 2011 Jan 13. Additional info
on importance of breakpoints can be obtained from 
* Nat Biotechnol. 2010 Jan;28(1):47-55. Epub 2009 Dec 27.
* Nature. 2011 Feb 3;470(7332):59-65.

The software runs on linux based systems.


1. Compilation
==============

$ make

or (without parallel support)

$ make OMP=no

which ever works


2. Running
==========

$ ./age_align file1.fa file2.fa

The input files must be in FASTA format and can contain multiple sequences.
The output is alignments for each pair of sequences with first sequence from
the first file and second sequence from the second file.

When aligning long read or assembled conting to a chromosome, useful options
are -coor1 and -coor2. These options allow specifying region(s) of the
inputed sequences, to be used in an alignment. For example, for a prediction of
a deletion in the region chr12:11396601-11436500 and assembled conting for
the allele containing this deletion, one may use the following command 

./age_align chr12.fa conting.fa -coor1=11395601-11437500

i.e. align conting to the the predicted region extended by 1 kb downstream and
upstream.

The penalty model used for determining the cost of insertions and deletions is 
the affine gap model. The gap penalty function G(i) is defined
as G(i) = go + (ge x i). Note that when there is a single gap then G(i) = go + ge

Help:
$ ./age_align


Options:

-indel			assume deletion or insertion (default)
-tdup			assume tandom duplication
-invl			assume inversion with conting (second sequence)
			spanning over left breakpoint
-invr			assume inversion with conting (second sequence)
			spanning over right breakpoint
-inv			assume inversion; tries alignment over the left and
			right breakpoints; report the best alignment
-coor1=start-end	align subsequence of first sequence defined by given
			coordinates
-coor2=start-end	align subsequence of second sequence defined by given
			coordinates
-revcom1		align reverse complement of first sequence
-revcom2		align reverse complement of second sequence
-both			align first sequence to second one and its reverse
			complement; report the best alignment
-match			score for nucleotide match
-mismatch		score for nucleotide mismatch
-go=value		gap open penalty, negative value
-ge=value		gap extend penalty, negative value
-allpos			always display boundary positions
-berg			align the sequences using Hirschberg's linear memory algorithm

Examples:

./age_align -coor1=20-2350    file1.fa file2.fa
./age_align -coor1=20-        file1.fa file2.fa
./age_align -coor2=-240       file1.fa file2.fa
./age_align -revcom1 -revcom2 file1.fa file2.fa
./age_align -both             file1.fa file2.fa
./age_align -inv  -both       file1.fa file2.fa
./age_align -tdup -both       file1.fa file2.fa





Please send your comments and suggestions to [email protected].