Amazing resource. Aside from being extremely useful for clinical decision making, it provides a great opportunity to better understand the regulatory architecture of the genome, given the tight link between loss-of-function intolerance and dosage sensitivity. (1/5) https://twitter.com/konradjk/status/1265663015564849152">https://twitter.com/konradjk/...
We can now ask if there are features specific to regulatory elements controlling highly LoF-intolerant genes. Not only is this extremely interesting in its own right, but such features can then be the building blocks for predictive models of LoF-intolerance (2/5)
These predictive models are still needed, because despite the sheer scale of gnomAD, for about 30% of coding genes constraint is uncertain. (3/5)
In https://www.biorxiv.org/content/10.1101/2020.02.15.936351v3,">https://www.biorxiv.org/content/1... we show that high CpG density at promoters is such a regulatory feature, and we use it to classify 1760 genes unascertained in gnomAD as highly LoF-intolerant or not. (4/5)
We think there is a lot more to be learned from this approach, especially regarding regulatory factor binding. See also https://www.sciencedirect.com/science/article/pii/S0002929720300124">https://www.sciencedirect.com/science/a... and https://www.nature.com/articles/s41467-018-04552-7">https://www.nature.com/articles/... (5/5)