Connectivity-based Bayesian nonparametric modeling of individual variability in autism

  • Awarded: 2019
  • Award Type: Pilot
  • Award #: 614379

Each individual with autism spectrum disorder (ASD) is unique, but there are patterns in individual variability. Understanding the structure of individual variability in ASD is critical: for research, because failing to model this variability may make it more difficult to identify biomarkers; and for treatment, because different subtypes of ASD may respond differently to interventions. In the psychiatric literature, two fundamentally different models of individual variability have existed. In the ‘clustering’ model, individuals are organized in types and subtypes (as seen for example in the Diagnostic and Statistical Manual of Mental Disorders [DSM]). In the ‘factor analysis’ model, variability is captured with a set of dimensions (as seen for example in the Research Domain Criteria [RDoC] framework developed by the National Institute of Mental Health).

In this project, Joshua Hartshorne and Stefano Anzellotti aim to study individual variability in univariate and multivariate connectivity across individuals with ASD. To do so, they will use recently developed Bayesian nonparametric techniques to analyze a large functional magnetic resonance imaging database from individuals with ASD and related neurodevelopmental disorders collected as part of the Simons Variation in Individuals Project (Simons VIP). Clustering and factor analysis models will be directly compared to identify an optimal model that explains individual variability and that will be useful both for predicting responsiveness to treatment and for tailoring individual-specific interventions.

Subscribe to our newsletter and receive SFARI funding announcements and news