Phylogenetic factorization of compositional data yields lineage-level associations in microbiome datasets.

Area of Science:

• Microbial Ecology
• Phylogenetics
• Bioinformatics

Background:

• Microbial community sequencing yields large datasets on relative abundances across various conditions.
• Associated phylogenetic data offer opportunities to study the impact of evolutionary history on microbial abundance.
• Existing methods may not fully capture the influence of phylogeny on community structure.

Purpose of the Study:

• To introduce a novel method, phylofactorization, for identifying phylogenetic drivers of microbial community composition.
• To demonstrate the utility of phylofactorization in analyzing microbial datasets from distinct environments.
• To showcase phylofactorization as a tool for dimensionality reduction, visualization, and ecological inference.

Main Methods:

• Development and application of the phylofactorization method.
• Re-analysis of existing microbial community datasets from human-associated and soil environments.
• Utilizing phylogenetic trees alongside relative abundance data.

Main Results:

• Phylofactorization effectively reduces the dimensionality of complex microbial community data.
• The method serves as an ordination and visualization tool, revealing underlying community structure.
• Phylofactorization successfully identifies specific phylogenetic branches associated with ecological traits and abundance patterns.

Conclusions:

• Phylofactorization is a powerful tool for integrating phylogenetic information into microbial community analysis.
• The method aids in understanding how evolutionary history shapes microbial ecology and trait distribution.
• Phylofactorization offers a framework for inferring the origins of functional traits along microbial phylogenies.