The impact of strain heterogeneity on genomic profiling accuracy in mixed-strain Escherichia coli metagenomes

  1. Vaishnavi Divya Shridar
  2. Stefan Engblom

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Metagenomic sequencing is increasingly used for surveillance of bacterial populations, enabling simultaneous characterisation of antimicrobial resistance, virulence, and epidemiologically relevant strain markers. However, the ability of commonly used bioinformatic tools to recover these features from heterogeneous populations remains incompletely understood.

Here, we developed a controlled benchmarking framework based on simulated single-species Escherichia coli metagenomes generated from genomes with PCR-determined annotations. Assembly-based and read-based tools were evaluated for antimicrobial resistance gene (ARG) detection, virulence gene (VG) detection, multilocus sequence typing (MLST), and serotype prediction. In addition, datasets with varying strain compositions were used to assess the effect of population complexity on strain-level characterisation.

We found that recovery of individual genetic determinants of interest (7 ARGs, 3 VGs) was consistently more successful than recovery of strain-level markers. ARG detection achieved high recovery rates across all evaluated tools (84.3–90.0%), while short read sequencing typer SRST2 achieved the highest virulence gene recovery rate (100.0%). In contrast, MLST and serotype prediction were strongly affected by strain composition. None of the evaluated MLST tools correctly assigned the dominant sequence type in the multi-strain log-normal dataset, despite successful partial recovery of MLST alleles, while complete serotype recovery remained uncommon across all methods. Decreasing the complexity of the simulated metagenome by increasing the abundance of the dominant strain improved both MLST and serotype prediction performance, while long-read analyses indicated that increased read length was insufficient to overcome the challenges associated with mixed-strain communities.

These findings demonstrate that performance estimates derived from gene detection do not necessarily translate to strain-level characterisation. While individual resistance and virulence determinants could often be recovered reliably, accurate reconstruction of sequence types and serotypes remained strongly constrained by strain heterogeneity. More broadly, the results highlight current limitations in resolving strain composition from metagenomic data and provide a framework for investigating the abundance thresholds and population structures required for reliable strain-resolved surveillance.

Contribution Statement

This study quantifies how strain heterogeneity influences recovery of clinically relevant genomic features from low-complexity Escherichia coli metagenomes. We show that gene-level characterisation remains largely robust, while strain-level markers such as MLST and serotype are substantially more sensitive to mixed-strain populations. These findings highlight strain heterogeneity as a major barrier to linking clinically relevant genes to their epidemiological background in metagenomic surveillance.

Article activity feed

  1. PREreview

    This Zenodo record is a permanently preserved version of a PREreview. You can view the complete PREreview at https://prereview.org/reviews/23897273.

    The impact of strain heterogeneity on genomic profiling accuracy in mixed-strain Escherichia coli metagenomes 

    Authors: Vaishnavi Divya Shridar and Stefan Engblom 

    Reviewer: Bali Schock

    Originally Published on bioRxiv

    1. Research Summary

    This study looks at how accurately different bioinformatics tools can identify antibiotic resistance genes (ARGs), virulence genes (VGs), multilocus sequence types (MLST), and serotypes in mixed-strain Escherichia coli metagenomes. The authors compare read-based and assembly-based approaches using simulated datasets with experimentally validated reference annotations. Through this process, the investigators found that individual resistance and virulence genes can be recovered with relatively high accuracy, while strain-level characterization is considerably more challenging. ARG recovery exceeded 84% across all evaluated tools, whereas none of the MLST tools correctly identified the dominant sequence type in the original mixed-strain datasets. The authors conclude that strain heterogeneity represents a significant limitation in metagenomic surveillance and that reliable gene detection does not necessarily indicate accurate strain-level characterization.

    This study addresses an important methodological challenge in expanding the use of metagenomic sequencing for antimicrobial resistance monitoring and wastewater-based epidemiology. Although resisting research has evaluated individual bioinformatic tools, fewer studies have systematically compared their performance across multiple clinically relevant genomic features under standardized conditions. Incorporating PCR-validated reference annotations and varying levels of strain complexity established a useful framework for identifying limitations in existing genomic profiling approaches. These approaches are relevant to public health surveillance, particularly in efforts to characterize circulating bacterial populations while monitoring epidemiologically important lineages.

    1. Examples and Evidence

    Major Comments: 

    • The primary concern is the exclusive use of simulated, single-species E. coli metagenomes rather than experimentally sequenced environmental samples. While this approach provides controlled conditions for evaluating strain heterogeneity, it does not fully capture the complexity of real wastewater samples, including diverse microbial communities, uneven sequencing coverage, and environmental contaminants. These differences have the potential to substantially influence bioinformatics tool performance. By incorporating real wastewater sequencing data, the study's external validity could be improved. Alternatively, the authors could more explicitly limit their conclusions to low-complexity, single-species metagenomic conditions.

    • This study evaluated antimicrobial resistance and virulence gene detection primarily through recovery rates, without assessing false-positive predictions or overall classification accuracy. While authors acknowledge that PCR-confirmed annotations do not establish gene absence, this approach limits conclusions about the reliability of these tools for surveillance applications. High gene recovery does not necessarily indicate high specificity, particularly in heterogeneous metagenomic samples. The authors could clarify this distinction, and consider incorporating reference datasets with confirmed negative annotations to permit a more comprehensive evaluation of tool performance.

    Minor Comments:

    • The authors generated only 10 metagenomes per genomic feature category, which may not adequately capture the variability in strain composition and genomic diversity encountered in real-world surveillance. This limited sample size reduces the statistical precision of performance estimates and makes it difficult to determine whether observed differences between bioinformatics tools are consistent across diverse conditions.

    Competing interests

    The authors declare that they have no competing interests.

    Use of Artificial Intelligence (AI)

    The authors declare that they did not use generative AI to come up with new ideas for their review.

  2. Version published to 10.64898/2026.10.19.760091 on bioRxiv