Using the SomaScan® Assay to identify protein biomarkers and create risk scores for Chronic Obstructive Pulmonary Disease (COPD) and Emphysema
ABSTRACT
COPD and emphysema are lung diseases related to chronic exposure to tobacco smoke or air pollution. The World Health Organization anticipates that COPD will be the number three cause of death worldwide by 2030; however, the majority of smokers and people exposed to air pollution do not develop COPD or emphysema. To better understand why only some people develop disease, we used SomaScan® v4.0 (4,776 human proteins) to identify plasma biomarkers of disease status as well as severity and progression in 6,017 subjects from the COPDGene cohort. Because we were able to identify hundreds of individual biomarkers, the majority of which explained only a small percentage of the phenotypes, we used penalized regression (LASSO) to develop protein risk scores, which typically include 200-300 of the most influential proteins. These protein risk scores explained as much as 40-50% variance of disease.
In this webinar you will
- Learn the basic clinical features and pathophysiology of COPD and emphysema
- Understand how the SomaScan Assay can be used to discover biomarkers
- Consider caveats on how to design good proteomic discovery studies and what potential biases should be considered
- Learn how to use penalized regression (LASSO) to develop protein risk scores
Russ Bowler, M.D., Ph.D.
Russell Bowler, M.D., Ph.D. obtained a B.S. in mathematical and computational sciences from Stanford University, a M.D. from the University of California at San Francisco (UCSF), and a Ph.D. in Cell and Developmental Biology from the University of Colorado (CU). He completed his internal medicine residency at UCSF and a pulmonary and critical care fellowship from CU.
The mission of Dr. Bowler’s lab is to understand the mechanisms of how cigarette smoke leads to the development of chronic obstructive pulmonary disease (COPD), the third leading cause of death in the United States. His team has generated genetic, genomic, and metabolic profiles on 10,000 subjects in the NIH sponsored COPDGene cohort and used these Omics data to identify novel diagnostic and therapeutic targets.
In 2020, Dr. Bowler and collaborators formed the SomaScan® Proteomics Consortium to promote collaboration among prospective cohort studies that follow participants for a range of outcomes and perform SomaScan profiling of individuals. The consortium facilitates an open exchange of ideas, knowledge, and results in order to accelerate the study of proteomics profiles associated with chronic disease phenotypes.
More webinars
WebinarBoutique Webinar Aptamers with protein-like side chains as a versatile tool for high-content proteomics
Proteins, encoded in 20,000 genes in humans, do much of the work in biology. Measuring proteins, which change in response to various perturbations and represent targets for almost all drugs, offers insights about the health status of an organism. Since proteins operate in complex networks rather than in isolation, measuring multiple proteins simultaneously offers richer insights compared to single protein measurements.
WebinarUsing Proteomics To Advance Understanding of Alzheimer’s Disease
Limited understanding due to its complex pathophysiology and lack of definitive biomarkers currently constrains the diagnosis and treatment of Alzheimer’s disease (AD). But new research is uncovering dynamic brain changes during Alzheimer’s progression, offering potential therapeutic targets. This webinar explores how proteomics and systems biology can be integrated to elucidate AD pathology.
WebinarPredictive modeling and reliable biomarker discovery in clinical omics studies
High-content omic technologies coupled machine learning methods have transformed the biomarker discovery process. However, the translation of computational results into scalable clinical biomarkers remains challenging. A rate-limiting step is the rigorous selection of reliable biomarker candidates among a host of biological features. Drawing examples from real-world clinical omics studies, I will introduce Stabl, a general machine learning framework that identifies a sparse, reliable set of biomarkers by integrating noise injection and a data-driven signal-to-noise threshold into multivariable predictive modeling.