Comparing genotyping chips to deep sequencing?

[dan]

Hi all,

I'm looking for any literature that compares SNP genotyping chips (such as infinium or illumina genotyping platforms) to 'deep sequencing' datasets such as WGS or RNASeq (and the subsequent SNP calling using SAMTools or similar).

I've seen a few human genome papers where the authors also include a SNP chip analysis (probably because it's so cheap compared to WGS), but I'm not sure of any specific studies or meta-studies.

Thanks for any pointers,
Dan.

[krobison]

Some genome centers (e.g. WashU) use the SNP chip info as a quality check, sample identity check & also to determine if they have gone deep enough.

I found this reference that did a direct comparison; I believe there are more

http://www.ncbi.nlm.nih.gov/pubmed/19420381

[dan]

Thanks very much for this link Kris. I'm looking to build a collection ;-)

All the best,
Dan.

[krobison]

You're welcome, but it's Lys-Glu-Ile-Thr-His not Lys-Arg-Ile-Ser :-)

[Michael.James.Clark]

We did so in the U87MG Decoded paper. Ran Illumina 1M-Duo and compared to our SNP calls from WGS on the SOLiD.

I think for these early studies it made sense, but we've reached the point where SNV calls from WGS are more accurate than the error rate of the SNP chips themselves. After that point, all we end up doing is saying, "Well, we did better than a SNP chip," which is kind of silly.

[dan]

Right, but there are different costs and requirements associated, so I'd like to know about the general characteristics so that the right choice can be made for the right study.

Thanks for the link, I'm happy to pull in all such papers ;-)

[dan]

You're welcome, but it's Lys-Glu-Ile-Thr-His not Lys-Arg-Ile-Ser :-)

Met-Ile-Asn-Thr-Glu-Arg!

I think I called you Kris a while back and for some reason it stuck.

Cheers,