The CFTR test you mentioned is a PCR test, the Illumina arrays are hybridization, correct?. If so, the two tests are not directly comparable in terms of error

The CFTR test you mentioned is a PCR test, the Illumina arrays are hybridization, correct?. If so, the two tests are not directly comparable in terms of error rates. I know that hybridization assays have good error rates, but that errors in calls do happen. Any time you test that many SNPs, errors happen. Do you also get the confidence measures for each call? But, I agree, this isn't really the issue. As I said below, I'm not trying to say that 23andMe shouldn't exist. I actually like them very much, and think that they've done a great job at getting the general public to understand how genetic variation affects your chances of developing a complex disease. I also think that they have an amazing data set that they can use to mine more genotype/phenotype interactions. I think that 23andMe are one of the good guys in this. But I do think that there needs to be a line between performing genotyping and giving out medical advice. I don't know where that line needs to be drawn. For example, the 'A/A' variant you used may be reported to be involved in diabetes in one paper, and then later explicitly found to not be involved in another. So, what do you tell the person who has the variant? What about after the second paper is published? Here's another example: a single deletion in CFTR (well, triad causing the loss of a phenylalanine) is responsible for 70-80% of cystic fibrosis cases. The remaining 20-30% are caused by hundreds of other variants. What if someone was concerned about being a carrier, and used one of the personal genetics services to find out. What if they had one of the rare variants and it wasn't on the array? They should have a medical professional help them interpret their results in order to make sure that they know the limitations of that testing. Final case: Let's say someone submits their DNA for genotyping and are told that they have the Huntington's variation. It would be grossly negligent to tell someone this without some degree of medical/genetic counseling follow-up.