Talk title: Genetic Architecture of Autism

Speaker: Yufeng Shen, PhD
Associate Professor, Department of Systems Biology
Associate Director, Columbia Genome Center
Columbia University

Abstract: 

Thorough understanding the genetic causes of autism is critical to improve clinical care. A common view of autism genetics is that common variants explain most of the phenotypic variance, while rare or de novo variants make up substantial attributable risk with large effect size. Recent studies estimated that there are about 1000 risk genes with large effects. However, only about 200 risk genes are known. Additionally, the partition of heritability between common and rare variants remains uncertain. While the SPARK project with 50,000 families will provide an opportunity to address these questions, a common challenge in genetic analysis is the uncertainty of the impact of missense variants. In this talk, I will go through our recent work on a new method (MisFit) for estimating the fitness impact of missense variants in human populations using machine learning, and then summarize new results from using MisFit in analysis of SPARK data, particularly, about the contribution of rare inherited missense variants to autism liability. 

Speaker Bio:

Yufeng Shen is an Associate Professor of Systems Biology and Biomedical Informatics at Columbia University. He received his B.Sc. in biochemistry from Peking University and his Ph.D. in computational biology from Baylor College of Medicine. At Baylor, he led the analysis of the first human genome sequenced by next-generation technologies. He currently directs research projects that integrate genomics data to predict the effect of genetic variation using statistical and machine learning methods and to apply genomics and computational biology in genetic studies of human diseases. His group developed CANOES (Backenroth et al 2014) for calling copy number variants from exome sequencing data, gMVP (Zhang et al 2022), and MisFit (Zhao et al 2023) for predicting pathogenicity and fitness effect of missense variants using machine learning. They discovered that epigenomic patterns in tissues under normal conditions are associated with risk genes of developmental disorders (Han et al 2018). In addition, his research led to the discovery of novel risk genes for congenital heart disease (Homsy et al 2015), congenital diaphragmatic hernia (Qi et al 2018), and autism (Zhou et al 2022).

Host:
Chao Xing, PhD