I think we can still be optimistic about several things:
- Larger data sets, combined with genetic simulations will allow us to narrow down the possible genetic architectures for the disease, as in this paper: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4158716/
- Eventually the answer will come down to gene x environment interactions. Even if we can’t fully account for the heritability of diabetes, we may be able to combine environmental risk factors with some genetic knowledge to identify certain people who have elevated risk or would be more responsive to certain therapies.
- Finding the genes where risk variants repeatedly show up will give us clues to disease mechanisms.
It’s a big task, and maybe T2 diabetes is on the extreme end. But since, genetics or not, large cohorts for disease studies are a good idea, and since the cost of genotyping drops ever lower, it seems like a good idea to keep pressing forward.