From Genes To Neural Systems: The Transcriptome As An Intermediate Phenotype

Jason Bohland

Junior Faculty Fellow, Hariri Institute for Computing
Department of Health Sciences, Sargent College
Boston University

Abstract: While it is clear that many neurological and neuropsychiatric disorders have strong genetic components, and that many aspects of brain function anatomy are heritable, relatively little is known about the large-scale molecular organization of the brain – that is, how genes drive brain structure and function. Using large publicly available transcriptomic data, I will describe several studies that aim to elucidate this organization, and demonstrate the value of multivariate gene expression profiles as signals related to regional differentiation in mouse and human brain. We use clustering, network analytic, and machine learning approaches to interrogate these data, revealing relatively strong separability of brain regions based on their expression profiles, even in the relatively uniformly organized cerebral cortex. These results are explained in terms of a simple model that describes gene expression at the mesoscopic scale as comprising a linear combination of distinct cell types. Finally, I will discuss the possible role that gene expression atlases may play in helping to elucidate the genetic underpinnings of heritable brain disorders.

Bio: Professor Jason Bohland joined the Department of Health Sciences at BU in 2009.  He previously completed postdoctoral work at Cold Spring Harbor Laboratory, helping to initiate the Brain Architecture Project, a large-scale informatics and experimental effort to assess neural connectivity patterns in the mouse brain. Jay completed his PhD in Cognitive and Neural Systems at Boston University, and also received MS and BS degrees in Electrical and Computer Engineering from the University of Cincinnati.  His research focuses on understanding the structural and functional architecture of neural systems in the human brain, with emphasis on those that support speech and language processes.  His laboratory uses an integrative approach that combines computational, informatics, and experimental brain imaging methods to elucidate these systems, and draws on a number of large public neuroscience databases.  An area of particular interest is in linking the molecular / genetic level of organization to the neural circuits / systems, and ultimately behavioral levels.