Suggestions for GWAS SOP

[lvclark]

I work for HPCBio, and Chris asked me to help out with the SOP for GWAS. I have made some edits and fleshed out some sections where I had the expertise to do so.

Here are a few other issues that I didn't know how to deal with:

• The section on random errors mentioned experiment-wide cutoffs. I mentioned the GenCall score in order to be more specific, but I don't know if the original author had anything else in mind.
• For the relatedness matrix a "genome xs" option for PLINK is mentioned. I couldn't find anything in the PLINK documentation about the "xs" portion of this.
• It is mentioned that in a homogenous population, one will see a single tight cluster in PCA, whereas in an admixed population, there will be a single loose, fuzzy cluster. Has this really been observed? I would expect a loose cluster for both, but with less variance explained on the first two axes for the homogenous population.
• Under the single-locus section, there was a "mixed model approaches... which..." bullet point. It seems incomplete but I don't know what the author intended to put there.

Anyway, let me know if I can be of any further help!

[cjfields]

@lvclark thanks, so far it looks great! We have a meeting on the SOPs set up Friday morning for several others working on this; I may ask at the next HPCBio meeting if you ran into any issues with the formatting or editing.

[cjfields]

Hi @shaze I'm going ahead with the merge on this. Should we maybe add JT to the repo as a reviewer?

[shaze]

I think a good idea