Attention:
Due to technical maintenance some tools will not be available.
See maintenance description for more information.

BiBiServ -
Bielefeld University Bioinformatic Service

OMA: Optimal Multiple Sequence Alignment

Authors: Torsten Will, Knut Reinert, Jens Stoye

Abstract

OMA is a library of C++-classes for the optimal alignment of sequences, built upon the Library of Efficient Data structures and Algorithms (LEDA). The parts of the library build a hierarchical, modular system, so that one can rearrange blocks or replace predefined modules by ones own classes. OMA is described in the extended abstract Combining Divide-and-Conquer, the A*-Algorithm and Successive Realignment Approach to Speed up Multiple Sequence Alignment which has been presented at the German Conference on Bioinformatics, GCB 99.

Examples

We have run OMA on a number of data sets from the literature.

	OMA runs *(modified: 2000/03/13)* on all data sets from reference1 of BAliBASE, and a comparison with MSA.
	Similar runs with and without use of the A*-Algorithm.
Early tests of OMA:
Another comparison of OMA with MSA (german). Comparison of the first iterative version of OMA and MSA (german). Different implementations of the trie used for storing the vertices and edges of the lattice (german). Early comparison of one dataset with and without A*.

Download OMA

You may download

	a precompiled version of oma for SUN Solaris (.gz, compiled with gnu egcs)
	Some information about this version of oma and some sample data (.tar.gz) that makes it easier for you to run some tests.
	the source code (.tar.gz) of the library OMA and the default implementation of the alignment program oma (and many intermediate test programs). Note that LEDA (version 2.7.3 or higher) is required to compile the source code;
	to recompute our above results you may download a version of oma (SUN Solaris, .tar.bz2, egcs) NOT using the A*-Algorithm (sorry, this was removed -- please contact the authors).

Licence agreement: You may use the library and the program for non-profit research and personal use freely. You may distribute the unchanged source code. If you make changes you must include the original source and notify the authors when redistributing. The authors are not responsible for any damage of data or hardware that the downloading or using of the program or library may cause.

If you use OMA in your own research, please cite:
Knut Reinert,Jens Stoye,Torsten Will:
The extended abstract Combining Divide-and-Conquer, the A*-Algorithm and Successive Realignment Approach to Speed up Multiple Sequence, (GCB 99).

Keep in touch

You can set yourself on a mailinglist so you are notified when changes of oma are made. Simply write an email to mailto:oma-announce-request@poempel.com?subject=subscribe with the subject "subscribe".
This is a standard mailinglist: You recieve postings to oma-announce@poempel.com. You can manage you subscription posting commands to mailto:oma-announce-request@poempel.com?subject=help and the subject "help".

Further References