You are here:
Decision tree-based method for integrating gene expression, demographic, and clinical data to determine disease endotypes
Citation:
Williams-DeVane, C., D. Reif, E. Hubal, P. Bushel, E. Hudgens, J. Gallagher, AND S. Edwards. Decision tree-based method for integrating gene expression, demographic, and clinical data to determine disease endotypes. BMC Systems Biology. BioMed Central Ltd, London, Uk, 7(1):119, (2013).
Impact/Purpose:
The understanding gained from the use of this case study will lead to better understanding of complex disease in general by comparison and application of more efficient methodologies.
Description:
Complex diseases are often difficult to diagnose, treat, and study due to the multi-factorial nature of the etiology. Significant challenges exist with regard to how to segregate indivdiuals into suitable subtypes of the disease. Here, we examine a range of methods for evaluating gene expression and clinical indicators of allergy and childhood asthma to inform the basic mechanistic underpinnings of disease etiology. We compared traditional methods such as Student's t-test and single data domain clustering as well as more complex, multi-data domain methods such as multi-step decision tree and modk-prototypes algorithm analysis strategies to determine the best method to segregate asthmatics. Traditional methods did not segregate asthmatics and non-asthmatics well, were difficult to interpret and provided sparse mechanistic insight. Methods that incorporate multiple domains of data performed better overall. The understanding gained from the use of this case study will lead to better understanding of complex disease in general by comparison and application of more efficient methodologies.