Thread by @RKCYuen

Our new study is now out in @nature! We identified repeat expansions as a substantial genetic contributor to #autism, in collaboration with 38 researchers across @SickKidsNews, @UofT, @illumina, @verilylifesci, @autismspeaks & more ⤵️ (1/9)

Link: nature.com/articles/s4158…

Thanks to advances in #genome sequencing, there have been improvements in clinical diagnosis & gene discovery for a variety of human disorders. Despite this, the number of genetic factors identified so far contribute less risk than expected - aka "missing heritability" (2/9)

This is also true for autism spectrum disorder (ASD). Despite >100 genes implicated in ASD, the genetic factors identified so far can only explain ~20% of cases 🤔 We thought that some answers to these genetically unsolved cases may lie in repetitive regions of the genome. (3/9)

Tandem repeat DNA makes up ~6% of the human genome. It can have complex structures, making it difficult to detect. That’s why ExpansionHunter Denovo (EHdn) was developed to search & find significant tandem repeat expansions in the genome ⤵️ genomebiology.biomedcentral.com/articles/10.11…

In this @nature study, we coupled EHdn with our novel analytic approach, involving a non-parametric outlier detection method, to identify repeat expansions in 17,231 genomes of families with autism, and controls, finding a variety of repeat expansions. (5/9) #SKResearch

We found:
📍 A total of 37,865 repeat motifs across 31,793 distinct regions of the human genome - that's ~1.2 different motifs per region!
📍 That more tandem repeat containing regions are found in GC-rich regions, in fragile site regions and in upstream & 5'UTR regions of genes.

📍 We discovered novel repeat expansions in genes responsible for nervous system development
📍 We found that repeat expansions were more prevalent among individuals with autism (23.3%) than those unaffected (20.7%), suggesting that repeat expansions contribute 2.6% risk to ASD

📍 Beyond implications for ASD, we found that there is wide sequence variability among repeat expansions than previously thought
📍 Overall, our approach to detecting repeat expansions across the genome is a model that can be applied to other complex disorders ☑️ (8/9)

We're grateful to all of the families for participating, and to the many researchers who made this study possible, including the few on Twitter: @MikeEberle, @egor_dolzhenko, @IanBackstrom, @EvdokiaAnagn, @ElsabbaghMayada & @mila_mirceta! (9/9)