I have fasta format. Can anyone say how to convert the fasta format to pir format.
Also, I have clustalw alignment format, then how to convert clustalw alignment format to PIR or .ali format.
I have fasta format. Can anyone say how to convert the fasta format to pir format.
Also, I have clustalw alignment format, then how to convert clustalw alignment format to PIR or .ali format.
I could help you out with specific problems, but I'm not sure I can do your search for you. Another option off the top of my head is readseq.
If you were to elaborate your first statement, specifically on how you think the output provided by EMBOSS does not conform to your requirement, perhaps it can be tweaked. If sequence conversion has been unsuccessful for so long, then either the approach or the requirement should be given a deeper look, IMO.
Here the link, says - 404 not found.
Actually, I need the O/P like this,
>P1;H2WHC9_CAEJA
sequence:: : : : :::-1.00:-1.00
MQRMIEMGGPHIEEIVDPPSPSNSVLQEADYRQNGPSRSRPKLIDQIRTLIRTPNFNKD------AVKM---GQA
ID------------LMELQIAEWVEEQIR------YGFTQEREDAIYQYRRKLRRQN---AKPRPLFSRKKFQM-
---FKKSVEEAEERYFEEREESRRREEARRAMSYSRGDI----SARDNRSKLHQSHT----QRNYGDSLDSDDEN
ERENGYQSYRPPPQRQQRLRSRSRSRSPMRSSYRHESPENSRRNASHQQTAQVRMRAGKNNVTKTKKWRPGQKAL
SEIRKYQNSTDLLIQKA-------PF----RRLVHQIIQEATGFDSGFRIRADAMSALQEAAEAFIVEMFEGSVL
ISNHAKRVTLMTADIQLYRRLCLRNL-------------------------------------------------
-*
Ah, that's because the URL added my parentheses to itself. A mere formatting issue, I apologize. The URL: http://iubio.bio.indiana.edu/soft/molbio/readseq/java/
EDIT on 14-Oct-2019: The URL issue has been corrected in the original comment.
Your requirement looks like aligned FASTA to me. Source on various alignment formats: http://emboss.sourceforge.net/docs/themes/AlignFormats.html
If you supply the right destination format specifier, EMBOSS should work. If you could paste a head
of your input, your actual command and the head
of your output, that would help.
Hi,
I'm suffering from the same problem as well, basically I'm using modeller and its functionality in my python scripts. Biopython deals with FASTA format, whereas to build a comparative model MODELLER uses PIR file to make use of structural information. I'm having a hard time to deal with this two formats. what I tried to do is first I obtain two sequences in FASTA format and then do aln.append(file = 'file.fasta', align_codes='all', alignment_format='FASTA')
then after that I did:
aln.write(file='5fd1_1fdx_output.fasta', alignment_format='FASTA')
aln.write(file='5fd1_1fdx_ouput.pir', alignment_format = 'PIR')
and used the latter (5fd1_1fdx_ouput.pir
) to build the model. but it's not working since I'm loosing information whenever I convert from FASTA to PIR.
so the input FASTA format file is(5fd1_1fdx_sequence.fasta
):
>5fd1
AFVVTDNCIKCKYTDCVEVCPVDCFYEGPNFLVIHPDECIDCALCEPECPAQAIFSEDEVPEDMQEFIQLNAELA
EVWPNITEKKDPLPDAEDWDGVKGKLQHLER
>1fdx
AYVINDSCIACGACKPECPVNIIQGS--IYAIDADSCIDCGSCASVCPVGAPNPED-----------------
-------------------------------
and the output file (5fd1_1fdx_ouput.pir
):
P1;5fd1 sequence:: : : : :::-1.00:-1.00 AFVVTDNCIKCKYTDCVEVCPVDCFYEGPNFLVIHPDECIDCALCEPECPAQAIFSEDEVPEDMQEFIQLNAELA EVWPNITEKKDPLPDAEDWDGVKGKLQHLER*
P1;1fdx sequence:: : : : :::-1.00:-1.00 AYVINDSCIACG--ACKPECPVN-IIQG-SIYAIDADSCIDCGSCASVCPVGA---------------------- -------------PNPED-------------* ```
Does anyone know how to convert it without loosing information?
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Jalview allows saving in PIR-format. You could try saving the alignment as PIR in Jalview?