YSPH Biostatistics Seminar: “Network analysis for understanding complex predictors in complex diseases: an example from proteomics in Alzheimer's Disease"

NOTE: BIS 525 students are required to attend in person. Others are invited to attend in person, but may also attend via Zoom.

SPEAKER: Meghan Short, PhD, Assistant Professor of Medicine, Tufts University School of Medicine; Faculty: Institute for Clinical Research and Health Policy Studies (ICRHPS); Faculty: Biostatistics, Epidemiology, and Research Design (BERD), Tufts Clinical and Translational Sciences Institute (CTSI)

TITLE: “Network analysis for understanding complex predictors in complex diseases: an example from proteomics in Alzheimer's Disease"

ABSTRACT: Weighted correlation network analysis (WGCNA) is a method for identifying groups of inter-correlated features within high-dimensional datasets such as gene expression and proteomic assays, which can then be related to disease outcomes. Such network-based approaches can be used to detect subtle but consistent variations in molecular profiles, that represent specific biological pathways or processes that may be implicated in health and disease. As an illustrative example, I will discuss a piece of ongoing research in Alzheimer’s Disease and Related Dementia (ADRD), where I use WGCNA to identify groups of correlated plasma proteins in a middle-aged, community-based cohort from the Framingham Heart Study, and measure their associations with brain MRI measures and incident dementia. The analysis, paired with an over-representation analysis using various annotation databases, identifies biological processes and pathways that may be involved in cognitive aging. I will compare the WGCNA method with complementary network-based approaches for association studies using ‘omics data, which have been applied to the plasma proteome in ADRD and have identified some associations in common with the WGCNA approach.