Fastq conversion using CASAVA

[tahamasoodi]

Hi,
I am using casava 1.8.2 for bcl to fastq conversion. I have few questions:

The command I use is:

/usr/local/bin/configureBclToFastq.pl --input-dir /usr/local/share/CASAVA-1.8.2/examples/Validation/110120_P20_0993_A805CKABXX/Data/Intensities/BaseCalls --output-dir Unaligned

I get the output and I can identify each sample separately without giving samplesheet.csv file

Questiotion 1) Is the samplesheet.csv necessary for the conversion? If we do not provide samplesheet, will it have any impact on the output?

Question 2) I am getting different numbers of fastq files for different samples like 10 files (5 for read 1 and 5 for read 2) for sample 1, 14 (7 for each read) for sample 2 and so on. Why is it like this? Is there any way that I can get a single file for each sample?

[GenoMax]

I get the output and I can identify each sample separately without giving samplesheet.csv file

Questiotion 1) Is the samplesheet.csv necessary for the conversion? If we do not provide samplesheet, will it have any impact on the output?

In this case the "Samplesheet" file may have been provided when the run was originally set up . Have you looked to see if there is a "Samplesheet" file in "Flowcell_ID/Data/Intensities/Basecalls" directory?

If you do not provide a sample sheet file then you may just get generic files that would have the lane number as file name. You will have to provide a sample sheet if you are multiplexing samples and want CASAVA to do the de-multiplexing.

Question 2) I am getting different numbers of fastq files for different samples like 10 files (5 for read 1 and 5 for read 2) for sample 1, 14 (7 for each read) for sample 2 and so on. Why is it like this? Is there any way that I can get a single file for each sample?

By default CASAVA produces fastq files in chunks of about 2 GB each to keep the downstream ELAND alignments from running out of memory. You can override the default by additng "--fastq-cluster-count 0" (that is a zero) option to your CASAVA command line to get single large files for read 1 and read 2.

[arkal]

If you do not provide a sample sheet file then you may just get generic files that would have the lane number as file name. You will have to provide a sample sheet if you are multiplexing samples and want CASAVA to do the de-multiplexing.

What he means is that if you haven't provided a samplesheet.csv (and if tehre wasn't one in your sequencing directory) then CASAVA will not register multiplexed lanes and all samples in a multiplexed lane will be assumed to be the same sample and treated accordingly as a single sample.

By default CASAVA produces fastq files in chunks of about 2 GB each to keep the downstream ELAND alignments from running out of memory. You can override the default by additng "--fastq-cluster-count 0" (that is a zero) option to your CASAVA command line to get single large files for read 1 and read 2.

^^ What he said

[sklages]

What he means is that if you haven't provided a samplesheet.csv (and if tehre wasn't one in your sequencing directory) then CASAVA will not register multiplexed lanes and all samples in a multiplexed lane will be assumed to be the same sample and treated accordingly as a single sample.

Yes, no demultiplexing with CASAVA. That's what GenoMax said.

^^ What he said

No, he was asking "Why is it like this?" and GenoMax was answering this question. He also pointed to "--fastq-cluster-count 0".