BCL to FASTQ conversation for paired-end RNA data

[seqmonkey]

Hi all,

I have paired end RNA seq data in bcl format that I want to convert to fastq format (using CASAVA 1.7/1.8).

I know that CASAVA doesn't do paired-end alignment for RNA data, so the plan is to align using a third-party tool such as BowTie.

The question is whether the data will be preserved in paired-end format after the conversion by default? If not, what steps are necessary to preserve the paired-end data?

[I am not sure if it makes any difference, but just as a note, the data will be demultiplexed as a separate step after FASTQ conversion because I am not able to do the bcl conversion and demultiplexing in one step (due to compatibility issues).]

Thanks!

[GenoMax]

Why are people suddenly getting BCL format data ? Are the sequencing facilities not doing due diligence for the data before it is released to users?

Seqmonkey: CASAVA v.1.8 will do the de-multiplexing and conversion to FASTQ as a single step. Ideally you should go back to your sequence provider and get them to do this for you. The paired end information is automatically preserved as a part of the conversion process. In fact you will get two separate files that contain the reads from the two ends. FYI: CASAVA software is not open source.

[senpeng]

we are looking into that too.
AND we used a different index ( four nucleotide instead of six ). thus casava keep saying "Clusters with unmatched barcodes for lane X". Any idea how to solve this?

[seqmonkey]

is this shen? i'm in the same lab as you

[GenoMax]

Are these custom barcodes? Did you use illumina multiplex protocol or is this a homebrew barcode protocol?

Originally Posted by senpeng

we are looking into that too.
AND we used a different index ( four nucleotide instead of six ). thus casava keep saying "Clusters with unmatched barcodes for lane X". Any idea how to solve this?

[senpeng]

Genomax,
Yes. I think the protocol is basically the same, we just used another kit instead of illumina standard truseq kit.

[GenoMax]

But were these barcodes part of the sequence itself or were they read separately as an independent read like Illumina's?

Originally Posted by senpeng

Genomax,
Yes. I think the protocol is basically the same, we just used another kit instead of illumina standard truseq kit.

[senpeng]

Thank you for your reply, GenoMax.
The barcode is not part of the sequence, it was used as index for demultiplexing reads. Illumina seems has its own set of barcodes.

Originally Posted by GenoMax

But were these barcodes part of the sequence itself or were they read separately as an independent read like Illumina's?

[GenoMax]

Then CASAVA should be able to demultiplex your reads. You will need to provide the right tags for your samples via the "SampleSheet.csv" file.

I do not want to sound discouraging but this is something that should be done by your sequence provider for you. It is not rocket science but it is not trivial to get it working right.

Originally Posted by senpeng

Thank you for your reply, GenoMax.
The barcode is not part of the sequence, it was used as index for demultiplexing reads. Illumina seems has its own set of barcodes.