Abstract
Patterns of sequencing coverage along a bacterial genome—summarized by a peak-to-trough ratio (PTR)—have been shown to accurately reflect microbial growth rates, revealing a new facet of microbial dynamics and host-microbe interactions. Here, we introduce CoPTR (Compute PTR): a tool for computing PTRs from complete reference genomes and assemblies. We show that CoPTR is more accurate than the current state-of-the-art, while also providing more PTR estimates overall. We further develop theory formalizing a biological interpretation for PTRs. Using a reference database of 2935 species, we applied CoPTR to a case-control study of 1304 metagenomic samples from 106 individuals with irritable bowel disease. We show that PTRs have high inter-individual variation, are only loosely correlated with relative abundances, and are associated with disease status. We conclude by demonstrating how PTRs can be combined with relative abundances and metabolomics to investigate their effect on the microbiome.
Introduction
Understanding the growth dynamics of microbial populations in the human gut is essential for deciphering host-microbe interactions. Traditionally, measuring growth rates required logistically challenging time-series sampling. However, the Peak-to-Trough Ratio (PTR) of sequencing coverage provides a "snapshot" alternative. Because bacterial DNA replication proceeds from a specific origin to a terminus, rapidly dividing populations exhibit higher sequencing coverage near the origin.
While promising, existing PTR estimation tools often struggle with Metagenome-Assembled Genomes (MAGs), which are typically fragmented into contigs of unknown order. To address these challenges, we developed CoPTR (Compute PTR), a robust computational framework designed to infer growth rates from both complete genomes and fragmented assemblies.
Methods: The CoPTR Framework
CoPTR implements two primary modules to handle different types of genomic data:
- CoPTR-Ref: Optimized for complete reference genomes, this module utilizes a maximum likelihood estimation (MLE) approach to fit a probabilistic model to read density. It incorporates rigorous filtering to remove noisy regions of anomalous coverage.
- CoPTR-Contig: Designed for fragmented "draft" assemblies, this module employs Poisson Principal Component Analysis (PCA) to analyze coverage patterns across multiple samples. By identifying shared fluctuations in coverage, CoPTR-Contig can reorder fragments based on their proximity to the replication origin, effectively reconstructing the genomic map required for PTR calculation.
Results and Validation
Validation using simulated datasets (E. coli, L. gasseri, and E. faecalis) demonstrated that CoPTR-Ref significantly outperforms previous methods like KoremPTR. Specifically, CoPTR-Ref achieved a Pearson correlation of $r > 0.95$ with ground truth growth rates using as few as 5,000 reads. In fragmented assemblies, CoPTR-Contig showed superior performance over existing tools such as GRiD and iRep, maintaining accuracy even with 50% genome completeness.
In a large-scale application to 1,304 metagenomic samples from patients with Irritable Bowel Disease (IBD), CoPTR revealed that microbial growth rates are highly personalized and only loosely correlated with relative abundance. For instance, Roseburia intestinalis—a beneficial butyrate producer—showed significantly altered growth rates in IBD patients that were not detectable through abundance metrics alone.
Discussion
CoPTR provides a powerful lens into the active dynamics of the microbiome. By decoupling growth from abundance, it reveals biological signals that traditional metagenomic analyses miss. While the tool is highly robust to assembly fragmentation and contamination, future work will focus on improving strain-level resolution and integrating PTR data with metabolomics to further explore the drivers of microbial growth in health and disease.
Summary updated