what's in solexa's control lane

[mingkunli]

I have been told that the phix genome used in solexa's control lane should be free of SNP, which means the reads should be exactly mapped to the phix genome if there is no sequencing eror, but as far as I know, Phix is a bacteria, which should have a high mutation rate, then how could they know no SNP in the control lane?

[mingkunli]

I found actually there are some SNP have been found, But if it's possible to have two different phix genome in the control lane? which will influence the concensus calling rather than sequencing error ?

So, I really don't know how the phix DNA in the control lane be prepared.

[ScottC]

PhiX174 is a bacteriophage. I guess there may be some snps developing over time, but as you mentioned, it probably depends how they maintain and prepare it. In the end, I don't think it matters all that much... some people don't use a control lane at all.

[mingkunli]

I used that data based on one assumption: only one genome there, I wonder to what extent it is true.

[SillyPoint]

We've had control lane PhiX samples from a couple of different sources, including Illumina and home-grown. We do see a few differences, which isn't surprising.

One batch featured a single position (1301) which reliably produced A and G calls about 50% each -- presumably indicating a mixture of two mutants in that batch.

My advice would be to align and SNP-call a couple of runs, figure out what you've actually got and (assuming consistent results of that test), use that for your reference from now on.

OTOH, as ScottC suggests, a few differences probably doesn't affect much.

--SP

[mingkunli]

That's exact what we found, 1301, with equivalent amount of A and G.
Actually, I am using this data to test my method to detect heteroplasmy, however, I am not sure whether the Phix data is clean.