Non-linear genetic regulation of the blood plasma proteome.

A recent preprint abstract discusses the need to understand how proteins are regulated by the genome and identifies a deep learning-based approach called EIR-auto-GP to study these effects. The study applied this method to a proteomics cohort of 48,594 individuals and identified 138 proteins that were regulated by non-linear effects, including genetic dominance and epistasis. The researchers also discovered a novel epistatic interaction between two loci and demonstrated dominance effects of a specific locus on plasma levels of certain receptors. The study concludes that large sample sizes are needed to uncover more complex genetic effects and that this approach can be applied to study other tissues or molecular QTLs. [Extracted from the article]