Through the sequencing efforts of the Simons Simplex Collection (SSC), Simons Foundation Powering Autism Research (SPARK) and Autism Sequencing Consortium (ASC), thousands of de novo coding mutations have been identified in people with autism spectrum disorder (ASD), the majority of which are de novo missense (dnMis) mutations. Thus far, however, dnMis mutations have not contributed much to our understanding of ASD, largely because their functional impact is unknown, unlike de novo loss-of-function mutations.
Haiyuan Yu’s group and others have found that current bioinformatics tools alone, e.g., combined annotation-dependent depletion (CADD) and polymorphism phenotyping v2 (PolyPhen-2), do not prioritize ASD risk variants effectively, and thus, additional information is needed. To imbue greater value in the dnMis mutation data, the current project brings together two groups with complementary skills and expertise. The Yu group has extensive expertise in high-throughput systems biology experiments, network analysis and machine learning. Kathryn Roeder’s group has a long history of developing innovative statistical frameworks to interpret genetic data, including widely used tools such as transmission and de novo association (TADA)1 and detecting association with networks (DAWN)2 to sift through de novo mutations and pinpoint genes involved in ASD risk. Supported by previous SFARI funding, Yu and Roeder successfully developed a novel integrated experimental-computational interactome perturbation framework to prioritize ASD dnMis mutations3,4.
Here, the researchers propose to generate the first cell-type-specific interactome networks for 269 high-confidence ASD risk genes and their 389 dnMis mutations in four relevant neural cell types using next-generation cross-linking mass spectrometry (XL-MS) technology. They plan to drastically improve their 3D interactome perturbation prediction pipeline (that can be applied to all missense mutations on all genes) by using the latest deep-learning models. They also plan to further develop the cell-type-specific DAWN method to integrate all experimental and computational results to determine subnetworks (i.e., communities) where unusual clustering of risk genes occur. Findings from these studies are expected to reveal key converging neurobiological pathways and sub-networks that modulate ASD risk.
- Interactome perturbation screen to identity damaging de novo missense mutations in autism
- Spatiotemporal and cell-type convergence to reveal autism neurobiology
- High-throughput autism variant functional testing using genetic interaction technologies in model systems
- Maximizing autism gene discovery by combining machine learning and single-cell expression data analyses