Conversion from bcl format to fastq files

[kjaja]

Hi,

I have sequencing data in raw format, the plan is to convert the .bcl files into fastq and demultiplexing using CSAVA. In order to do that I need to provide a sample sheet file. Does anyone know if I need to create this file and what files to use in order to do that.

Thanks for your help

[clariet]

Yes, you need to create it. A csv file with the following header:

FCID,Lane,SampleID,SampleRef,Index,Description,Control,Recipe,Operator


Sorry, I got it confused with demultiplexing. For bcl conversion, you don't need any sample sheet file. Just run this python script (setupBclToQseq.py) provided by Illumina.

Clariet

[GenoMax]

Kjaja,

I find it very surprising that you have been given raw data without conversion (and possibly internal QC) by your sequence provider. Do you have access to CASAVA software? Processing with CASAVA does have a learning curve. It may be more work than necessary if you are planning to do this for just one data set.

[simonandrews]

Sorry, I got it confused with demultiplexing. For bcl conversion, you don't need any sample sheet file. Just run this python script (setupBclToQseq.py) provided by Illumina.

That depends on the version of Casava you're using. For 1.8 you need a sample sheet even if you're not demultiplexing since it uses the sample and project names to construct the output file and directory names.

I agree with other comments though - if this is just a one off I'd go back to your sequencing provider and ask them to do the conversion for you. You don't want to have to get your head around Casava just for one sample.

[kjaja]

thank you all for your help, but what does the "Recipe" in the sample sheet stands for, I guess how do I get this information. also, can I tell from my files what the sample ID's are?

[simonandrews]

To just run the extraction the only fields you actually need to fill in are, lane, sample name, contol and project name (and barcode index if you need to demultiplex). You can leave all of the rest blank and it won't make any difference to the output. All of the other fields are only there to keep a record of what happened in your sample, the pipeline doesn't actually use them in any way.

The sample IDs are just arbitary names you've give to your samples. If you don't care about these you can just use Sample_1 Sample_2 etc.

The recipe is the set of instructions used to control the sequencer and normally just corresponds to the type of run you've done.