Approximately 100 risk genes for autism spectrum disorder (ASD) have emerged from studies focused on identifying germline de novo mutations (GDMs) within the exome. However, the genetic risk from several classes of genetic variation — in particular, mosaic mutations — has yet to be fully explored due to technical challenges or underpowered studies.
Postzygotic mosaic mutations (PMMs), aka somatic mutations, can occur throughout development. Importantly, mutations that occur during early embryonic development can be distributed to all three germ layers and can even be transmitted to the next generation if they are present in gonadal cells. Thus, PMMs can be potentially as damaging as GDMs but their detection and recurrence risk are complex.
Brian J. O’Roak and his colleagues previously developed, with support from an earlier SFARI award, a sophisticated approach for discovering and validating PMMs using whole-exome sequencing data from the Simons Simplex Collection1. Based on these findings, they hypothesize that PMMs are major genetic risk factors for ASD and that they will be enriched in distinct biological pathways compared to GDMs.
To test this hypothesis, O’Roak’s team, together with collaborator Jacob Michaelson at the University of Iowa, aim to leverage the current SPARK cohort data from more than 9,000 families to fully elucidate the role of PMMs in ASD risk and discover the biological mechanisms unique to germline and mosaic mutations.
Firstly, they plan to characterize mosaic mutations in the SPARK cohort in order to validate previous findings of mosaic mutation burden and to determine rates of mutation in the early embryo.
Secondly, they plan to jointly analyze germline and mosaic mutations from the SPARK cohort, using a combined mutational framework, in order to identify novel ASD risk genes.
Thirdly, they plan to leverage single-cell genomics data and a new machine learning based tool (forecASD)2 to discover the biological mechanisms, developmental time points and cell types that are disrupted by these distinct classes of de novo mutations.
Overall, the methods and tools developed here will increase our understanding of the diverse genetic and biological mechanisms that underlie autism risk.
- Somatic mosaicism in autism spectrum disorders
- Investigating the role of somatic mutations in autism
- A somatic mechanism for autism phenotypic heterogeneity
- Maximizing autism gene discovery by combining machine learning and single-cell expression data analyses
- Spatiotemporal and cell-type convergence to reveal autism neurobiology