Post

I'm excited to share that our work studying gene dosage response curves (GDRCs) is now out in Cell Genomics (@cellpress.bsky.social). www.cell.com/cell-genomic... [1/n]

With amazing collaborators @courtsmithrun.bsky.social, @huishengzhu.bsky.social, @tamigj.bsky.social, and my mentors @jeffspence.github.io and @jkpritch.bsky.social. [2/n]

We observed that the average effect of gene deletions (LoF) and gene duplications was positively correlated across traits. We use the GDRC to conceptualize what types of gene dosage architectures may explain this pattern. [3/n]

We find that this pattern is explained by trait buffering, which is the propensity for GDRCs to be buffered against modifying the trait in one particular direction over the other. [4/n]

We propose a simple model of an upstream "dysregulation" phenotype that may explain why GDRCs organize themselves in this manner. [5/n]

I think GDRCs are a conceptually pleasing way to tie together various estimates of gene effect that are used in the field (e.g., burden tests, TWAS). [6/n]

In the paper, we introduced new quantities related to GDRCs: monotonicity and trait buffering. These are somewhat challenging to interpret, so I have a short blog post explaining them: nikhilmilind.dev/posts/interp... [7/n]

We also did some extensive work to figure out variants and individuals to include in our burden tests. The key takeaway is that we can be smart with our variant filtering, and we can include everyone in the UKB! nikhilmilind.dev/posts/synony... [8/n]