details of *.csfasta creation

[d17]

Hi,

I am trying to figure out exactly where a *.csfasta file resulting from an ABI SOLiD run comes from. In particular, I am wondering whether the csfasta files are generated from image files alone.

As best I can tell, the raw image files from the SOLiD machine are combined and clustering is used to generate per-base data for each bead on the flow cell. This data is then run through some simple filters and all reads that are full length and have a color call for each position are output to the csfasta file. Is this accurate?

One reason I am curious about this is that I know that the *_sequence.txt files generated by the Illumina GA do NOT come from image files alone --- as I understand it, the standard analysis pipeline includes calibration to a reference genome. I am wondering whether this might also be the case for the standard SOLiD pipeline.

[drio]

There is no filtering happening for SOLiD. The equivalent for .csfasta + .qual files in solexa would be the qseq files.

[snetmcom]

the system does not write the csfasta straight from the images. There are several intermediate files that are used to dump data.

[d17]

the system does not write the csfasta straight from the images. There are several intermediate files that are used to dump data.

Thanks for the info. I am still a little confused, though: are the intermediate files generated from only the image files? And then the .csfasta + .qual are the result of filtering for reads that meet some quality threshold?

Or does any other information besides the image files (such as a reference genome) go into the ultimate production of the .csfasta file?

[snetmcom]

so as drio said, there is no filtering of reads for the creation of the csfasta. There is no chastity filter to throw out reads. This is why your mapping % looks lower on SOLiD while more reads still map.