2021/03/06 by Pratheepa Jeganathan, Susan Holmes, Jeganathan, Pratheepa +1
Biochemistry, Genetics and Molecular Biology · Computer Science · #Gut microbiota and health #Bacterial Identification and Susceptibility Testing #Bayesian Methods and Mixture Models
paper · pdf · doi:10.48550/arxiv.2103.04198
High throughput sequencing (HTS)-based technology enables identifying and\nquantifying non-culturable microbial organisms in all environments. Microbial\nsequences have enhanced our understanding of the human microbiome, the soil and\nplant environment, and the marine environment. All molecular microbial data\npose statistical challenges due to contamination sequences from reagents, batch\neffects, unequal sampling, and undetected taxa. Technical biases and\nheteroscedasticity have the strongest effects, but different strains across\nsubjects and environments also make direct differential abundance testing\nunwieldy. We provide an introduction to a few statistical tools that can\novercome some of these difficulties and demonstrate those tools on an example.\nWe show how standard statistical methods, such as simple hierarchical mixture\nand topic models, can facilitate inferences on latent microbial communities. We\nalso review some nonparametric Bayesian approaches that combine visualization\nand uncertainty quantification. The intersection of molecular microbial biology\nand statistics is an exciting new venue. Finally, we list some of the important\nopen problems that would benefit from more careful statistical method\ndevelopment.\n