Limits...
Does a barcoding gap exist in prokaryotes? Evidences from species delimitation in cyanobacteria.

Eckert EM, Fontaneto D, Coci M, Callieri C - Life (Basel) (2014)

Bottom Line: Nevertheless, the application of predetermined threshold in genetic distances to identify units of diversity (Operative Taxonomic Units, OTUs) may provide biased results.The application of a tool developed for animal DNA taxonomy, the Automatic Barcode Gap Detector (ABGD), revealed that a barcoding gap could actually be found in almost half of the datasets that we tested.The identification of units of diversity through this method provided results that were not compatible with those obtained with the identification of OTUs with threshold of similarity in genetic distances of 97% or 99%.

View Article: PubMed Central - PubMed

Affiliation: Microbial Ecology Group, Institute of Ecosystem Study, National Research Council, Largo Tonolli 50, 28922 Verbania, Italy. e.eckert@ise.cnr.it.

ABSTRACT
The amount of information that is available on 16S rRNA sequences for prokaryotes thanks to high-throughput sequencing could allow a better understanding of diversity. Nevertheless, the application of predetermined threshold in genetic distances to identify units of diversity (Operative Taxonomic Units, OTUs) may provide biased results. Here we tests for the existence of a barcoding gap in several groups of Cyanobacteria, defining units of diversity according to clear differences between within-species and among-species genetic distances in 16S rRNA. The application of a tool developed for animal DNA taxonomy, the Automatic Barcode Gap Detector (ABGD), revealed that a barcoding gap could actually be found in almost half of the datasets that we tested. The identification of units of diversity through this method provided results that were not compatible with those obtained with the identification of OTUs with threshold of similarity in genetic distances of 97% or 99%. The main message of our results is a call for caution in the estimate of diversity from 16S sequences only, given that different subjective choices in the method to delimit units could provide different results.

No MeSH data available.


Related in: MedlinePlus

Plot of the distribution of pairwise genetic distances in a hypothetical group of organisms of different species. All the genetic distances that belong to pairwise comparisons of organisms within the same species fall in the bars on the left, in this case below 2.5%; all the distances that belong to pairwise comparisons between organisms of different species fall in the bars on the right, in this case between 4.5% and 9.5%; no intermediate distances exist between the two distributions, defining a dataset-specific barcoding gap ranging from 2.5% to 4.5%. In this case, a 97% threshold would provide reliable units of diversity, whereas a 99% threshold would overestimate the actual biological diversity.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4390840&req=5

life-05-00050-f001: Plot of the distribution of pairwise genetic distances in a hypothetical group of organisms of different species. All the genetic distances that belong to pairwise comparisons of organisms within the same species fall in the bars on the left, in this case below 2.5%; all the distances that belong to pairwise comparisons between organisms of different species fall in the bars on the right, in this case between 4.5% and 9.5%; no intermediate distances exist between the two distributions, defining a dataset-specific barcoding gap ranging from 2.5% to 4.5%. In this case, a 97% threshold would provide reliable units of diversity, whereas a 99% threshold would overestimate the actual biological diversity.

Mentions: Enormous progress happened in prokaryote taxonomy in the recent years [8,12,24]. However, the data produced with novel methodologies, such as next generation sequencing, often requires a high throughput taxonomic classification of sequences, such as fixed threshold to identify OTUs. This kind of fixed numeric classification can always only be a vague approximation to the actual structure of relatedness of organisms. Surprisingly, only few tests have been performed to assess whether a barcoding gap, similar to the one commonly found in DNA barcoding markers in animals [25], actually exists in prokaryotes (e.g., [26,27]). The existence of a barcoding gap would mean that genetic distances within each species are short, genetic distances between species are long, and no intermediate distances are present (Figure 1). A test on the existence of such barcoding gap in prokaryotes will help researchers deciding which threshold to use and whether separate units can be actually found using 16S. The untested application of fixed thresholds (97% or 99%) in genetic distances will always produce separate units, regardless of whether they are meaningful or not; on the other hand the existence of a barcoding gap will show that such units are biological reality.


Does a barcoding gap exist in prokaryotes? Evidences from species delimitation in cyanobacteria.

Eckert EM, Fontaneto D, Coci M, Callieri C - Life (Basel) (2014)

Plot of the distribution of pairwise genetic distances in a hypothetical group of organisms of different species. All the genetic distances that belong to pairwise comparisons of organisms within the same species fall in the bars on the left, in this case below 2.5%; all the distances that belong to pairwise comparisons between organisms of different species fall in the bars on the right, in this case between 4.5% and 9.5%; no intermediate distances exist between the two distributions, defining a dataset-specific barcoding gap ranging from 2.5% to 4.5%. In this case, a 97% threshold would provide reliable units of diversity, whereas a 99% threshold would overestimate the actual biological diversity.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4390840&req=5

life-05-00050-f001: Plot of the distribution of pairwise genetic distances in a hypothetical group of organisms of different species. All the genetic distances that belong to pairwise comparisons of organisms within the same species fall in the bars on the left, in this case below 2.5%; all the distances that belong to pairwise comparisons between organisms of different species fall in the bars on the right, in this case between 4.5% and 9.5%; no intermediate distances exist between the two distributions, defining a dataset-specific barcoding gap ranging from 2.5% to 4.5%. In this case, a 97% threshold would provide reliable units of diversity, whereas a 99% threshold would overestimate the actual biological diversity.
Mentions: Enormous progress happened in prokaryote taxonomy in the recent years [8,12,24]. However, the data produced with novel methodologies, such as next generation sequencing, often requires a high throughput taxonomic classification of sequences, such as fixed threshold to identify OTUs. This kind of fixed numeric classification can always only be a vague approximation to the actual structure of relatedness of organisms. Surprisingly, only few tests have been performed to assess whether a barcoding gap, similar to the one commonly found in DNA barcoding markers in animals [25], actually exists in prokaryotes (e.g., [26,27]). The existence of a barcoding gap would mean that genetic distances within each species are short, genetic distances between species are long, and no intermediate distances are present (Figure 1). A test on the existence of such barcoding gap in prokaryotes will help researchers deciding which threshold to use and whether separate units can be actually found using 16S. The untested application of fixed thresholds (97% or 99%) in genetic distances will always produce separate units, regardless of whether they are meaningful or not; on the other hand the existence of a barcoding gap will show that such units are biological reality.

Bottom Line: Nevertheless, the application of predetermined threshold in genetic distances to identify units of diversity (Operative Taxonomic Units, OTUs) may provide biased results.The application of a tool developed for animal DNA taxonomy, the Automatic Barcode Gap Detector (ABGD), revealed that a barcoding gap could actually be found in almost half of the datasets that we tested.The identification of units of diversity through this method provided results that were not compatible with those obtained with the identification of OTUs with threshold of similarity in genetic distances of 97% or 99%.

View Article: PubMed Central - PubMed

Affiliation: Microbial Ecology Group, Institute of Ecosystem Study, National Research Council, Largo Tonolli 50, 28922 Verbania, Italy. e.eckert@ise.cnr.it.

ABSTRACT
The amount of information that is available on 16S rRNA sequences for prokaryotes thanks to high-throughput sequencing could allow a better understanding of diversity. Nevertheless, the application of predetermined threshold in genetic distances to identify units of diversity (Operative Taxonomic Units, OTUs) may provide biased results. Here we tests for the existence of a barcoding gap in several groups of Cyanobacteria, defining units of diversity according to clear differences between within-species and among-species genetic distances in 16S rRNA. The application of a tool developed for animal DNA taxonomy, the Automatic Barcode Gap Detector (ABGD), revealed that a barcoding gap could actually be found in almost half of the datasets that we tested. The identification of units of diversity through this method provided results that were not compatible with those obtained with the identification of OTUs with threshold of similarity in genetic distances of 97% or 99%. The main message of our results is a call for caution in the estimate of diversity from 16S sequences only, given that different subjective choices in the method to delimit units could provide different results.

No MeSH data available.


Related in: MedlinePlus