Hi, we have a small set of 12 Illumina Human Omni SNP arrays we've run. I'm just getting into this after dealing with gene expression for a long time, and SNP array analysis is a different world. We have a few samples from disease and a few from control, all independent samples (no families). I know this is far too few samples to actually get any sort of disease association, but I need to try to do what I can to extract something of potential utility.
I've got "finalreport" files with "Allele 1- TOP" and "Allele 2- TOP" along with the log R ratio and the B allele frequency. I tracked down a PFB file and ran PennCNV. The case/control analysis function in PennCNV is very limited though. I don't have access to Genomestudio- the arrays were run/processed by a collaborator.
Any suggestions for other tools to try to run? Does it make sense to try PLINK even without family info? I've seen suggestions to just do a Fisher exact test on the genotype calls across samples to see if there's anything significant between the groups, but there have to be more possibilities than just that, even with this very limited sample size.
Any suggestions welcome.
