BowlofBabelFish
While I agree with most of your posts, I have to pick you up on your response to the abstract Wakame posted in support of genetic evidence for 'transness' (for want of a better word)
That paper is a fishing excercise. I’ve done loads of array work and it fails all the usual statistical checks.
It's not an array, it was exome sequencing, and as it's only an abstract (presumably for a talk or poster) I don't see how you can pass judgement on the statistical methodology without seeing more details.
An array is effectively a chip, with cut up bits of dna on. You get the RNA readout from a sample (which is all the actively genes that are being read and used) and turn it into DNA. Then you wash that over the chip and examine it. So a gene which is being highly expressed will have a stronger signal.
Exome analysis, which is what is described in that abstract, looks for variants or mutations in the germline and has nothing to do with RNA. The first grant I was principal investigator on did something similar.
Obviously you can look at different gene variants too - there are thousands of these variants between any two random people though so what you can’t do is just say look at ‘anything that comes up.’ Because if you fling enough spaghetti at the wall some will stick. 21000 genes, and a confidence level of 95% gives you roughly a thousand genes that will pop up as positive for chance just by chance. Ooops.
Obviously don't disagree with this.
So you have to be careful with arrays. There are very sophisticated analysis programs that allow you to reduce the false positives but by its nature an array is a fishing expedition. And that’s fine, as long as you then go into the source material and verify those gene changes and what they mean. So I’d then get actual tissue samples and look at whthervthat gene really was being expressed at a higher level. That paper didn't do that either - they stopped at the fishing.
One person's fishing expedition is another's unbiased, agnostic approach...
Yes, the data would need to be confirmed, but I don't think anyone is claiming it's anything but preliminary.
They also looked a sex steroid metabolism genes in a population that could be treated with sex steroids. It’s a bit like saying that body builders on anabolic asteroids have changes and conveniently ignoring the facts the drugs they are taking cause them.
Again, there are not enough details in the abstract to criticise the methodology. Yes, it's a potential problem, but it may or may not have been addressed.
As a geneticist (as I am also) I would be interested in your opinion on this paper www.ncbi.nlm.nih.gov/pubmed/22146048. I can only see the abstract (as I am no longer in academia and don’t have institutional access), but it seems fairly convincing to me that there is a genetic component to gender dysphoria. Maybe you can get the whole paper?