A robust, sensitive phylogenetic method enables gene-level metagenomic analyses

Area of Science:

• Microbiome research
• Computational biology
• Genomics

Background:

• Current microbiome research aims to link microbial gene function to specific mechanisms, moving beyond simple taxonomic associations.
• Existing methods often identify marker genes with limited functional relevance, especially with differential abundance data.
• Phylogenetic regression, while useful for prevalence changes, shows inflated false positives with differential abundance data.

Purpose of the Study:

• To develop a robust statistical method for linking microbial gene function to differential abundance data.
• To address the limitations of phylogenetic regression in analyzing metagenomic datasets.
• To enable the generation of new molecular hypotheses from microbiome case-control studies.

Main Methods:

• Developed and validated a non-parametric method called "robust permutration" for differential abundance analysis.
• Compared robust permutration against phylogenetic regression and other phylogenetic comparative methods using simulations.
• Applied robust permutration to a human liver cirrhosis case-control metagenomic study.

Main Results:

• Phylogenetic regression exhibits an anti-conservative bias with inflated false positives in differential abundance analysis.
• Robust permutration demonstrates superior power while appropriately controlling the false positive rate in simulations.
• The method identified a link between *Lachnospiraceae* abundance in liver cirrhosis and an uncharacterized iron-sulfur transcription factor.

Conclusions:

• Robust permutration is a powerful and reliable method for inferring microbial gene function from differential abundance data.
• This method facilitates the generation of novel molecular hypotheses directly from metagenomic studies.
• The findings highlight a potential role for a specific microbial gene in liver cirrhosis pathogenesis.