Hacker News Reader

Comment by kayo_20211030 | Hacker News Reader

kayo_20211030·

I spend no time considering genetic variations, or genetic correlations, I deal with what's in front of me. This study is _almost_ backcasting; it might improve the model, but, not the outcome.

Maybe the study is fine and valuable, and maybe it'll lead to something. Maybe? But, it does nothing in the present. Not in the here and now.

View on HN

If we can find specific subtypes it would be extremely helpful. You can’t expect everything to be presented only with a complete solution.

Exactly, lumping in all autism into the same overall category when there could be very different underlying biological mechanisms would potentially block progress.

For example, a medication that could work wonders on one subtype by affecting a biological mechanism unique to that subtype could be found not to meet clinical standards because it didn't work on the other subtypes or even make them worse.

In other words, it's a confounding variable which needs to be discovered and characterized after which it could play significant role in advancing treatments and understanding.

That is a fair point, or two. My frustration with studies of this type is that they provide very little guidance with respect to how to prevent a condition, how to fix a condition, or how to manage a condition. Nor, do they give me much hope that even a long way down the road will they do so. My focus, because of a family situation, is management. Whether there's one type, four types, or more than a thousand types is, to me, immaterial. Even if this study was a complete genetic explanation of the mechanism, it wouldn't help, because I don't work with genes, I work with gene expressions, manifestations and symptoms.

Reply on news.ycombinator.com