Extract paired end reads from sff file.

[ojy]

Hi,

I'm working on a genome assembly project from 454 paired end reads which are in sff format. I have already tried Newbler which works with sff files directly, however now I want to try other things as well, so I need to have reeds in a user friendly format like fasta.

I used sff_extract whith -l and -c options to remove linkers and clip the ends. I decided to verify the resulting fasta file and assembled it with Velvet as single end data, however the results were very strange: N50 was about 15-30 bps. This is very strange given that the coverage is decent (~30) and single end reads assembly by Newbler was good. So I assume that maybe the extraction process went wrong and something else has to be clipped from the sequences? Any suggestions how to do it? Or did I do something wrong?

I was thinking to map the resulting fasta reads to Newbler contigs to see what is wrong with the reads. But I have not worked with alignment tools yet. I would appreciate any suggestions on how to align reads to contigs (preferably with visualization).

Thanks

[flxlex]

You could also run a newbler assembly with the '-tr' flag set, this will give you all trimmed reads as a separate output file. The paired end reads will be split in two (with linker removed, and clipped). Paired reads can be recognized by the '_left' and '_right' at the end of their names.

Perhaps this helps?

[ojy]

flxlex, thanks! This is exactly what I was looking for.

[zhangju]

What is the command line? I got GS2.6 here, and I have not run with command lines yet. In the path, I have list of commands - newbler, newAssembly, gsAssembly, runProject .... Which one I should use for extracting paired end reads from sff file?

Thanks,

Justin

You could also run a newbler assembly with the '-tr' flag set, this will give you all trimmed reads as a separate output file. The paired end reads will be split in two (with linker removed, and clipped). Paired reads can be recognized by the '_left' and '_right' at the end of their names.

Perhaps this helps?

[flxlex]

runAssembly -o some_name -tr path/to/454reads.sff