Error using sff_extract.

[Mona]

Hi,
i m trying to extract fasta, qual and xml files from .sff file using sff_extract.

I have just all the provided commands for extraction, Following commands have been used.

sff_extract filename.sff

sff_extract -o filename.sff

$ sff_extract -s seq.fasta -q qual.fasta -x anci.xml my_454_file.sff

With all the above commands, i get the same error message: "struct.error: unpack requires a string argument of length 1600"
How i can remove this error?

[maubp]

This suggests either the SFF file is damaged (e.g. truncated due to a problem downloading - try redownloading it), or doesn't quite follow the standard sff_extract is expecting. Is this an original unmodified SFF file from your sequencing center?

Do you have the (Linux only) Roche tools to test with?

[Mona]

yes this is the original file from the sequencing centre.
and i dnt have the roche tools to test with.

[maubp]

Getting the Roche "Newbler" tools is quite easy, assuming you have a suitable Linux machine to run them on. See:

http://seqanswers.com/forums/showthread.php?t=114

However, in the short term as mentioned above, I would strongly suggest you try re-downloading the SFF file from the sequencing center, or at least confirm you got the whole file by checking the filesize.

P.S. Seeing the whole error message might help.

[Mona]

Is it possible to get error in sff_extract when there is no vector used in the 454 sequencing.
As i know that the centre from which my data is coming they are not using any vector for sequencing. But A and B adapters are used. so is it still compulsory to screen the sequences and extract the fasta, qual and xml files from the sff file. and second thing do i have to perform the sequence masking for the tag sequence after the extraction of these three files? or only one of the above two steps is required.

[maubp]

Do you have an example SFF file with no vector used in the 454 sequencing that breaks sff_extract that you could share? I'd like to use it to check the SFF support in Biopython.

(Regarding the rest of your question, it probably depends on what you are going to do next - if you are using MIRA then it might be easier to ask on the mira_talk mailing list)

[Mona]

Yes i have the file that breaks sff_extract.
Can u send me ur email id?

[maubp]

After chatting to Mona privately, it does sound like the original problem was a partial SFF file. The full file works fine with sff_extract

[Mona]

Yes exactly.
Thnx maubp!!