Medical Image Analysis of Brain Images and Structural Networks for Alzheimer's Disease (SJH); Context-Specific Nested Effects Models (YS)

Tuesday, March 7, 2017 -
4:00pm to 5:00pm
Biotechnology Center Auditorium, 425 Henry Mall

Speaker Name: 

Seong Jae Hwang; Yuriy Sverchkov

Speaker Institution: 

Computer Sciences (SJH); Biostatistics & Medical Informatics (YS), UW-Madison

Cookies: 

No

Description: 

Medical Image Analysis of Brain Images and Structural Networks for Alzheimer's disease (SJH):
Understanding neurodegenerative diseases such as Alzheimer's disease often involves medical image analysis of various types of brain images (i.e., MRI and DTI) to infer important differences or patterns among subjects. In this talk, I will present some basic statistical analysis procedures of brain images to provide a brief overview of common tasks in medical imaging analysis. Next, I will describe a more specific characterization of brains as connectivity networks where the nodes correspond to distinct brain regions and the edges encode the strength of the connection, and show how graph-based techniques could be used to analyze and answer the main scientific question of interest: characterizing the structural integrity of brain networks from healthy to diseased individuals.

Context-Specific Nested Effects Models (YS)
Advances in systems biology have made clear the importance of network models for capturing knowledge about complex relationships in gene regulation, metabolism, and cellular signaling. A common approach to uncovering biological networks involves performing perturbations on elements of the network, such as gene knockdown experiments, and measuring how the perturbation affects some reporter of the process under study. We develop context-specific nested effects models (CSNEMs), an approach to inferring such networks that generalizes nested effect models (NEMs). The main contribution is that CSNEMs explicitly model the participation of a gene in multiple pathways, meaning that a gene can appear in multiple places in the network, each of which constitutes a different context for the gene, and corresponds to a different pathway. Biologically, the representation of regulators in multiple pathways may indicate that these regulators have distinct roles in different cellular compartments or life cycle phases. We present an evaluation of the method on simulated data as well as an application to data studying sodium chloride stress response in Saccharomyces cerevisiae, where we do find that the method identifies genes which participate in multiple pathways.