To Dereplicate or Not To Dereplicate?

Area of Science:

• Microbial genomics
• Bioinformatics
• Evolutionary biology

Background:

• Metagenome-assembled genomes (MAGs) are essential for studying microbial life.
• Existing tools may produce MAGs not representing natural populations.
• Redundancy among MAGs from independent datasets poses challenges.

Purpose of the Study:

• To investigate the issue of highly similar MAGs from independent datasets.
• To evaluate the impact of redundancy removal on MAG quality and gene content.
• To explore methods for analyzing closely related populations without dereplication.

Main Methods:

• Analysis of published MAG datasets for redundancy (>99% ANI).
• Comparison of different software packages and parameters for MAG dereplication.
• Assessment of gene loss in closely related genomes after redundancy removal.

Main Results:

• A significant fraction of published MAGs share >99% average nucleotide identity (ANI).
• Dereplication software and parameters vary widely in the number of MAGs removed.
• Removal of closely related genomes results in the loss of population-specific auxiliary genes.

Conclusions:

• Redundancy in MAG datasets is common and impacts downstream analyses.
• Dereplication strategies need careful consideration to avoid losing biological information.
• Alternative approaches can capture strain-specific dynamics without complete genome dereplication.