Erratum: Functional metagenomic profiling of nine biomes

Elizabeth A. Dinsdale, Robert A. Edwards, Dana Hall, Florent Angly, Mya Breitbart, Jennifer M. Brulc, Mike Furlan, Christelle Desnues, Matthew Haynes, Linlin Li, Lauren D. McDaniel, Mary Ann Moran, William Nelson, Christína Nilsson, Robert Olson, John H. Paul, Beltran Rodriguez Brito, Yijun Ruan, Brandon K. Swan, Rick Stevens, David L. Valentine, Rebecca Vega Thurber, Linda Wegley, Bryan A. White, Forest RohwerNature 2008 · DOI · OpenAlex

Abstract

This corrigendum addresses a mathematical error in the 2008 study "Functional metagenomic profiling of nine biomes," where functional diversity and evenness were originally calculated using base-10 logarithms instead of natural logarithms. While absolute values and evenness metrics have been updated, the relative diversity rankings between biomes remain consistent with the original findings, providing a precise baseline for global metagenomic comparisons.

Introduction

In 2008, Dinsdale et al. published a landmark study that shifted the focus from taxonomic profiling ("who is there") to functional profiling ("what are they doing") across nine distinct biomes. By sequencing the metagenomes of microbial and viral communities, the study sought to determine if different environments shared a common functional core or possessed unique metabolic signatures. However, a technical error was identified in the calculation of diversity indices: the Shannon diversity index ($H'$) and resulting evenness values were computed using a base-10 logarithm instead of the standard natural logarithm. This corrigendum corrects these discrepancies to ensure comparability with other ecological literature.

Methods

The original study utilized high-throughput metagenomic sequencing to sample DNA from microbes and viruses across nine biomes, mapping sequences to the SEED functional hierarchy. In this correction, the authors recalculated:

  1. Shannon Diversity ($H'$): Recomputed using the natural log ($\ln$). The maximum theoretical diversity ($H'_{max}$) for this dataset is corrected to 6.483.
  2. Evenness ($E$): Recalculated using the corrected $H'$ values and a corrected formula to fix an original error in the calculation pipeline.

The nine biomes analyzed include aquatic (Marine, Freshwater, Hypersaline), host-associated (Fish, Terrestrial animals, Mosquitoes), and specialized/benthic (Subterranean, Coral reefs, Microbialites) environments.

Results

The corrected diversity and evenness metrics for both microbial and viral fractions are summarized in Table 1.

Table 1: Corrected functional diversity (natural log) and evenness for microbial and viral communities across nine biomes.

Key results from the recalculation include:

  • Highest Microbial Diversity: Observed in Fish ($5.634 \pm 0.007$) and Freshwater ($5.595 \pm 0.015$).
  • Highest Viral Diversity: Observed in Fish ($5.615 \pm 0.001$) and Mosquitoes ($5.514 \pm 0.060$).
  • Evenness Extremes: Terrestrial animals exhibited the highest viral evenness ($0.965 \pm 0.009$), whereas Coral microbes showed the lowest microbial evenness ($0.748 \pm 0.040$).

Discussion

The primary finding of this corrigendum is the robustness of the original biological conclusions. Despite the shift in absolute values due to the change in log-base, the relative levels of diversity between biomes remained unchanged. This suggests that functional signals in metagenomic data are resilient to minor mathematical scaling errors.

The corrected evenness values (ranging from 0.74 to 0.96) indicate that functional roles within these biomes are well-distributed rather than dominated by a few housekeeping functions. Furthermore, the divergence between viral and microbial diversity in certain biomes remains significant; for example, in Coral biomes, viral functional diversity ($H' = 5.271$) is notably higher than microbial diversity ($H' = 3.991$), suggesting specialized viral roles in reef ecosystems.

While modern sequencing technology has advanced significantly since 2008, this correction provides an accurate historical baseline. Future work continues to address the challenge of "dark matter" in metagenomics—sequences with unknown functions—which remains a limitation in fully capturing ecosystem metabolic potential.

Summary updated

BibTeX
@inproceedings{dinsdale2008erratum,
  author = {Elizabeth A. Dinsdale and Robert A. Edwards and Dana Hall and Florent Angly and Mya Breitbart and Jennifer M. Brulc and Mike Furlan and Christelle Desnues and Matthew Haynes and Linlin Li and Lauren D. McDaniel and Mary Ann Moran and William Nelson and Christína Nilsson and Robert Olson and John H. Paul and Beltran Rodriguez Brito and Yijun Ruan and Brandon K. Swan and Rick Stevens and David L. Valentine and Rebecca Vega Thurber and Linda Wegley and Bryan A. White and Forest Rohwer},
  title = {{Erratum: Functional metagenomic profiling of nine biomes}},
  booktitle = {Nature},
  year = {2008},
  doi = {10.1038/nature07346},
}

Loading related content...

Figure