[GIRI]
Browse Search Repeat Masking Download Submit Repbase Reports Education

Alignment package


Name
Last modified
Size
alnpack-1.0.tar.gz
07-Mar-2016 05:19
371K
alnpack-2.0.tar.gz
07-Mar-2016 05:19
355K
alnpack-2.1.tar.gz
07-Mar-2016 05:19
360K

Description

Alnpack is a set of programs and a C++ library for pairwise local alignment of DNA or protein sequences using Smith-Waterman algorithm.

The package includes an optimized version of the algorithm originally developed by Michael Farrar. A Gotoh's backtracing was added to it in this package to produce alignments.

To take advantage of the optimized version of the algorithm computer's CPU must support SSE2 instructions. Most modern x86 processors have this support. Also your C++ compiler must support SSE2 intrinsics.

Another alignment implementation in the package uses Smith-Waterman algorithm with Gotoh's backtracing and logarithmic gap penalties. This version is about 3 time slower than the one mentioned above, but it does not rely on streaming SIMD instructions and provides benefits of logarithmic gap penalties which are more natural for biological data alignment.

The alignment programs depend on boost C++ libraries, which is a standard package in most Linux distributions. You will need boost development package in order to compile alnpack.

References

Alnpack was written by Oleksiy Kohany

Alnpack includes in part alignment software by Michael Farrar:

Farrar M.
Striped Smith-Waterman speeds database searches six times over other SIMD implementations.
Bioinformatics. 2007 Jan 15;23(2):156-61. Epub 2006 Nov 16.

Changelog

2.1
-Redesigned base classes for storing sequences and alignments.
-Fixed bug when programs were attempting to read non-existent protein alignment matrix file.
-Slightly improved performance.

2.0
-Alignment library completely rewritten with object-oriented approach in mind without impacting performance.
-Significantly improved performance of Monardo-Gotoh alignment algorithm implementation.

1.0
-Initial version


© 2001-2018 - Genetic Information Research Institute