Extrapolation of Urn Models via Poissonization: Accurate Measurements of the Microbial Unknown
2011

Accurate Predictions of Microbial Diversity Using Poissonization

publication Evidence: high

Author Information

Author(s): Lladser Manuel E., Gouet Raúl, Reeder Jens

Primary Institution: Department of Applied Mathematics, University of Colorado, Boulder, Colorado, United States of America

Hypothesis

Can we accurately predict the fraction of microbial species that remain unsampled in a given environment?

Conclusion

The study presents a new statistical method that provides accurate predictions for the fraction of unsampled microbial species in various environments.

Supporting Evidence

  • The Embedding algorithm provides conditionally unbiased predictors for unsampled microbial species.
  • Predictions were tested against simulated environments based on human-gut and -hand microbiota datasets.
  • The method allows for accurate predictions even with small sample sizes.

Takeaway

This study helps scientists understand how many different types of tiny living things, like bacteria, are in a sample, even if they haven't found them all yet.

Methodology

The study uses a new statistical method called the Embedding algorithm, which models microbial samples as draws from an urn with an unknown composition.

Limitations

The method may yield inconclusive predictions when sample sizes are fixed and may not apply to all datasets.

Digital Object Identifier (DOI)

10.1371/journal.pone.0021105

Want to read the original?

Access the complete publication on the publisher's website

View Original Publication