New autism genetics--looking through the lens of liability threshold model

•

Two great papers published today in @NatureGenet dive deep into the genetics of autism and autism related traits and paint a big picture of the underlying genetic architecture. 1. nature.com/articles/s41588-022-01064-5 2. nature.com/articles/s41588-022-01072-5

The first paper by Antaki et al emphasizes the differential contributions of various types of genetic variants--de novo/inherited, rare/common--to autism risk. Here is a detailed thread from the senior author. twitter.com/sebatlab/status/1532456270661402650?s=20&t=bc1X8nutuzRSe99kHkaEZg

The second paper by Warrier et al (nature.com/articles/s41588-022-01072-5) follows a similar theme as Antaki et al but with a special focus on the rare and common variant contributions to autism associated features. Here is my past thread on this paper. twitter.com/doctorveera/status/1424143355261427713?s=20&t=bc1X8nutuzRSe99kHkaEZg

One main take away from both the papers that is worth highlighting is that genetic contributions to autism risk are best appreciated through the lens of liability threshold model

The disease liability comes either mainly from rare variants with minor contributions from common variants or vice versa. It doesn't mean that it's not possible to have similar contributions from both. It's just that current sample sizes are underpowered to capture them.

This knowledge is crucial to efficiently identify the genetic risk factors in a given group of individuals with autism or to study the well known parental age and sex effects on autism risk.

Either searching for common variants in individuals who derive their disease liability predominantly from rare de novo variants (typically individuals presenting with co-morbid intellectual disability)

or searching rare variants in individuals who derive their disease liability predominantly from common variants (typically individuals with high functioning autism) might not be productive (at least in small to moderate sample sizes).