In article <JUN.96Mar29114706 at crick.nih.go.jp>, jun at nih.go.jp (Jun Ishikawa) writes:
> Thank you for your comment.
>> In article <4jdjbm$mn7 at ratatosk.uio.no> rodrigol at biotek.uio.no (Rodrigo Lopez) writes:
>>> ClustalW can write GCG's MSF format. Please check the following WWW page:
>> I know that. But I have already aligned 8000 bp x 20 seqs.,
> which took me about a week to get the result. So, I would like
> to convert it to MSF.
>
Clustal W can read a clustal-format alignment as input; if you do NOTHING
(i.e. do not do the alignment again), the alignment is preserved and you
can get it out again in a new format.
You might have to make sure that the version you have has enough sequence
length capacity (i.e. set MAXLEN in clustalw.h)
1) input the alignment (e.g. blah.aln) at the "input sequences from disc" menu
item.
2) go to the multiple alignment menu (but do NOT do any alignments :-).
3) select "output format options"
4) turn OFF clustal format output and turn ON msf format.
5) select "output the alignment now" and you should get blah.msf)
The above only applies to clustalw (not clustalv which is earlier).
Best of luck,
Des Higgins
>> readseq is another program that can convert multiple sequence alignment files
>> from/to a variety of formats including GCG's MSF, clustal and phylip.
>> I tried to use readseq, but failed:
>> BEFORE (filename is temp.aln):
> CLUSTAL V multiple sequence alignment
>> hpv01 GTTAACTACCATCATTCATTATTCTAGTTACAACAAGAACCTAGGAGTTAT-----ATGC
> hpv02 ATAATGTATAACTATAA-------TCCTT-TATTTAAAAATAGGGTGTGACCGAAAACGG
> hpv03 TCTAACTATAATTATAAATAACAATGCACATAATAAAAAGTAGGGAGTAACCGAAAACGG
>> AFTER (filename is temp.msf):
> temp.msf MSF: 29096 Type: N January 01, 1776 12:00 Check: 7280 ..
>> Name: temp.aln Len: 29096 Check: 7280 Weight: 1.00
>> //
>> temp.aln CLUSTALVmu ltiplesequ encealignm enthpvGTTA ACTACCATCA
>> temp.aln TTCATTATTC TAGTTACAAC AAGAACCTAG GAGTTAT--- --ATGChpvA
>> temp.aln TAATGTATAA CTATAA---- ---TCCTT-T ATTTAAAAAT AGGGTGTGAC
>> temp.aln CGAAAACGGh pvTCTAACTA TAATTATAAA TAACAATGCA CATAATAAAA
>> Readseq, really, can covert Clustal to MSF?
> Or, new version exists? # I used to version 1Feb93.
>> Regards,
>> # Sorry for my poor English.
> --
> Jun Ishikawa, Ph.D. National Institute of Health
> (jun at nih.go.jp) Dept. of Bioactive Molecules
> Tokyo, JAPAN
