,
5 tweets,
1 min read
Read on Twitter
N=15 patients with autism spectrum disorder. Fifteen. One-five. And how many data points are trawled through to find some differences? <checks details for precise figure>... a gazillion.
This for a disorder that is famously heterogeneous, clinically and genetically. And if any of the observed differences are actually real (which I wouldn't bet on), why take them as causes rather than consequences?
And what is the underlying hypothesis of these types of studies? Should we expect the convergence underlying similar symptom profiles to be happening at the level of gene expression in particular cells? Why would it be?
Autism isn't cancer. It doesn't reflect an altered state of gene expression. It reflects an altered state of neural systems. It may have genetic (and thus molecular) origins but that's no reason to expect it have proximal molecular mechanisms
If genomics is the answer, what's the question? wiringthebrain.com/2018/12/if-gen…
