(56) Multiple Sequence Alignment by Parallel Simulated Annealing

	Machine:     PIM
	Environment: PIMOS, SIMPOS
	Language:    KL1, ESP
	Source Code: 0.4 MB
	Documents:   Manual (English)

Overview

A Program for Protein Sequence Analysis.

Configuration

Generator:
Generating a process network for temperature parallel SA
SA Executer:
Calculating SA of each temperature

Function

This program aligns several protein sequences of weak homology in high quality by temperature parallel simulated annealing.

Multiple alignment is powerful method to analyse the protein or DNA sequences, by which we can recognise the conserved parts considerd to be significant. This is an example of an alignment.

 
    --HEKLLHPGIQKTTKLF-GET--YY
    -LHQ-LTHLSFSKMKALLERSHSPYY
    ELHS-FTHCG---QTAL--TLQ----
      ^    ^        ^
      |    |        |
Here`-'s(gaps) are inserted in order for each column to have the same or similar amino acids. Multiple alignment problem may be considered as a combinatorial optimization problem and is solved by simulated annealing method.

The conventional simulated annealing requires temperature scheduling which significantly depends on the problem, data and condition. Temperature parallel simulated annealing invented by Kimura(ICOT), however, distributes SA's of different temperatures into different Processors and calculates in parallel. This improvement make it needless to schedule the temperature and the optimal result is assured if the calculation is carried out enough. Thus this program produces very high qualitied alignments compared with that by a biologist.

FTP

www-admin@icot.or.jp