>confusing GWASing with variant calling >2018 >being a bird
Reminder: you may think GWASs of ~1 million people are large and extraordinary. But the fast remains that the human genome is huge, and if we want to find rare variants of small effect size, we will need to include perhaps 100 million people in some cases. https://www.sciencedirect.com/science/article/pii/S0002929717302409 …pic.twitter.com/gAYHeBm6yu
-
-
This Tweet is unavailable.
-
-
This Tweet is unavailable.
-
This Tweet is unavailable.
-
This Tweet is unavailable.
- End of conversation
-
-
-
The SSGAC FAQ is claiming they think only 20% of EA variance will ever be detectable by GWAS. What do you think?
-
These studies usually underclaim. That's how the politics work to get out in these journals I suspect. H2 for education length/degree is 65% in Norway when modeled properly, so the upper bound for genomic prediction is r = sqrt(.65) = .80.https://osf.io/preprints/socarxiv/fby2t/ …
End of conversation
New conversation -
-
-
Hm... From that power analysis description, that sounds like it's assuming individual variant testing. Isn't this ignoring more powerful analyses? eg once you've explained 80% of height variance with the lasso on WGS, rest of the rare variants should be far easier to detect.
-
Lots of assumptions, yes. Correlated variants increase the SE, and various other multivariate things like suppression effects, interactions, make it really tricky to power-estimate.
End of conversation
New conversation -
-
-
How many satellites does that buy you? How important are these satellites?
End of conversation
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.