Tuesday, December 10 2019
1:30pm - 3:30pm
Masters Presentation
Methodological Considerations for the Analysis of Longitudinal Metabolomic Data

Increased efficiency in collecting and analyzing the low-weight chemical compounds involved in metabolism (metabolomics) has enabled researchers to track the metabolite composition of individuals across time (a longitudinal design). The resultant data tends to be correlated and high-dimensional with hundreds to thousands of compounds at each time point, requiring specialized statistical techniques. In this paper we discuss the utility and construction of two such techniques: the ANOVA-simultaneous component analysis (ASCA; Smilde et al 2005) and the integrative conditional autoregressive horseshoe prior (iCARH; Jendoubi 2018). The conceptual and mathematical structures of both methods are covered and key similarities and differences between the two approaches are discussed. Model assumptions, interpretability, and complexity are given precedence for comparisons. We then provide potential areas for improvement to both methods and discuss future directions for longitudinal metabolomics data analysis
Speaker:Nicholas Weaver
Location:LW 700

Email Address:  
   Sign up to be notified by email before this event takes place.

Download as iCalendar