Outlier Detection for Minor Compositional Variations in Taxonomic Abundance Data

Abstract

To understand the activities of complex microbial communities in various natural environments and living organisms, we need to capture the compositional changes in their taxonomic abundance. Here, we propose a new computational framework to detect compositional changes in microorganisms, including minor bacteria. This framework is designed to statistically assess relative variations in taxonomic abundance. By using this approach, we detected compositional changes in the human gut microbiome that might be associated with short-term human dietary changes. Our approach can shed light on the compositional changes of minor microorganisms that are easily overlooked.

Keywords

• microbiome
• taxonomic abundance
• outlier detection
• computational approach
• relative variation