Bayesian Functional Data Analysis Over Dependent Regions and Its Application for Identification of Differentially Methylated Regions

• DOI: 10.1111/biom.13902
• Publication Date: 2023-07-22T02:53:27Z
• Authors: Suvo Chatterjee; Shrabanti Chowdhury; Duchwan Ryu; Sanjib Basu
• Keywords: Identification; Smoothing

We consider a Bayesian functional data analysis for observations measured as extremely long sequences. Splitting the sequence into several small windows with manageable lengths, may not be independent especially when they are neighboring each other. propose to utilize smoothing splines estimate individual patterns within window and establish transition models parameters involved in address dependence structure between windows. The difference of groups individuals at can evaluated by Bayes factor based on Markov Chain Monte Carlo samples analysis. In this paper, we examine proposed method through simulation studies apply it identify differentially methylated genetic regions TCGA lung adenocarcinoma data.