XSQ converter

[paolo.kunder]

I have a PE run, 6 lane, 12 barcoded samples.
ICS SOLiD software generated 6 XSQ file ( one per lane).

Now I will use XSQ_Tools to split each XSQ file in 12 indexed XSQ.

I will end up with 72 XSQ indexed files (6 lanes x 12 samples).

Is there a way to combine the 6 XSQ files prior to split them in order to generate at the end only 12 XSQ indexed files?

May I use merge Linux command to combine the 6 XSQ files or I may lose information?

Thanks,
Paolo

[genehunter]

Xsq

No, you cannot combine them using linux commands. The files are binary.
You may want to use LifeScope or HDF (format) APIs.

[colindaven]

We have been doing alignments of all and recombining - merging - the SAM files with Picards MergeSamFiles.

This is admittedly not a great solution !

Let us know if you have any better ideas.

[paolo.kunder]

Finally I converted my indexed XSQ files in csfasta (and QV.qual) files with XSQconverter and merged each individual csfasta (and QV.qual) with cat function,
information seems to be maintained,
paolo

[colindaven]

Yeah, did that too. Bioscope and Lifescope wouldn't use all the reads for alignment, but just the first set.
Not sure why that was.

NovoalignCS doesn't seem to pick up on the different read sets and uses all reads.

Colin

[kevleb]

Finally I converted my indexed XSQ files in csfasta (and QV.qual) files with XSQconverter and merged each individual csfasta (and QV.qual) with cat function,
information seems to be maintained,
paolo

When you cat files, you should obtain a final file with 6 times the same bead_id for different sequence comming from each of your 6 lanes. I'm not sure it won't be a problem then in the future bam file.
For external pipelines using .csfasta, instead of cat file, i open each .csfasta (and QV.qual) and update the panel number:
lane 1 : 1 to 708
lane 2 : 709 to 1416
lane 3 ...

In case of you use lifescope you should not do anything, import .xsq in lifescope and create reads set directly in lifescope. You can merge the .bam files from each lane with samtools after mapping. There you will have the bead_id issues in your final bam file. It seems not to be problem in case of only visualization in genome browser but it certainly depends on what you plan to do with the merge .bam file.

kevin.