Bacterial species: genomes, names and clinical consequences
Hosts Lee Katz and Nabil-Fareed Alikhan discuss bacterial taxonomy with Leighton Pritchard and Conor Meehan. They examine how biochemical tests, 16S sequences and whole-genome comparisons contribute to species definitions, and why unculturable organisms challenge the naming system. Clinical examples show the tension between describing evolutionary relationships and giving people useful, recognisable names.
Listen to Episode 67: Bacterial Taxonomy: what is a species, what is a strain? on SoundCloud
In this episode
What makes a bacterial species?
Pritchard traces his interest in taxonomy to diagnostic work involving plant pathogens and human pathogens associated with plants. That need also prompted him to write pyani. Meehan comes to taxonomy from an applied, clinical point of view: when does changing a name improve biological accuracy, and when does it confuse the people using it?
Neither offers a universal species definition. Meehan points out that reproductive definitions do not cover all multicellular organisms either. Pritchard looks for a barrier or discontinuity in gene flow, while acknowledging that his perspective comes mainly from genome sequences. Both stress that genomic classifications have largely been fitted back onto groups already recognised through biochemical and physical characteristics.
From biochemical tests to ANI
The species-description process discussed combines biochemical and phenotypic evidence with genetic comparisons. Tests might distinguish growth on particular substrates, colony appearance, morphology or motility. The panel describes a requirement for a culturable strain deposited in two collections, with ATCC given as one example.
Meehan traces an approximate historical correspondence between 97% 16S sequence similarity, 70% DNA–DNA hybridisation and 95% average nucleotide identity, or ANI. These comparisons connect newer measurements to previously recognised species rather than starting classification from scratch. ANI compares matching regions of two genomes; the discussion later returns to why focusing on shared regions matters.
At the time of recording, the guests say that formal instructions still call for a 16S tree, but reviewers increasingly expect genome sequences, ANI calculations and digital DNA–DNA hybridisation.
Genomes without cultures
Metagenome-assembled genomes, or MAGs, expose a limitation of that framework. A genome can be compared computationally, but without the organism in culture researchers cannot supply the same phenotypic and biochemical evidence or deposit a living type strain.
Pritchard says proposals to accept genome sequences as type material had been rejected by the ICSP. He presents this as a policy decision, not an inability to compare genomes. Meehan argues that some microbes will not be culturable on their own, leaving substantial diversity outside the formal naming scheme. Both expect change; Meehan warns that excluding genome-only organisms could push researchers towards competing taxonomies.
Names are not evolutionary histories
Meehan distinguishes a validly published name from the name a community currently prefers. Older names remain available and persist in papers, so adopting a replacement does not erase the historical record. Pritchard explains why this matters for literature searches: a paper describing an organism under its former genus may not be found by searching only its new name.
NCBI is the practical reference point for many users, although a host notes its disclaimer about not being the official taxonomy. The examples include taxonomic identifiers 1773 for Mycobacterium tuberculosis and 666 for Vibrio cholerae. The panel separates taxonomy as a description of evolutionary relationships from nomenclature as the labels attached to organisms.
When clinical names remain useful
E. coli and Shigella illustrate the tension between genomic grouping and clinical communication. The panel describes Shigella as embedded within E. coli, while stressing that recognising it separately remains important for clinical decisions.
Meehan describes Mycobacterium tuberculosis and Mycobacterium bovis as the same species, while Mycobacterium africanum comprises two lineages separated on the tree by animal-associated lineages. Retaining different names can lead clinicians to assume that fundamentally different treatment is required, while abandoning familiar names can also confuse readers.
He also contrasts Mycobacterium marinum, associated with disease in fish, with Mycobacterium ulcerans, associated with severe skin ulceration in humans. He says ANI groups them as the same species because it examines shared genomic regions, but their different disease consequences make clear identification important.
What deserves a lineage name?
Meehan recalls a discussion involving Ford Doolittle at Dalhousie University in which two researchers argued about lineages while using different definitions: taxonomic groupings versus descent through parents and grandparents. They were not talking about the same thing at all, and Meehan says the experience taught him that taxonomy is a useful term but should not be the be-all and end-all.
He also cautions against publishing a new lineage name for every clade on a tree. A branching pattern does not necessarily establish a biologically or clinically important distinction. He sees this particularly in virology, where diversity can be divided rapidly into named groups even when those groups do not do anything different.
Highlights
- 00:03:44 — Why even multicellular organisms resist a universal species definition
- 00:05:10 — Biochemical species descriptions, culture deposits and the historical role of 16S
- 00:06:55 — Fitting genetic measurements onto historically recognised species
- 00:08:37 — A discussion with Ford Doolittle reveals competing meanings of lineage
- 00:12:17 — ANI, its approximate 95% boundary and expectations for genome evidence
- 00:13:30 — Why MAGs cannot satisfy the same phenotypic requirements as cultured organisms
- 00:18:30 — Historical names in the literature and the distinction between taxonomy and nomenclature
- 00:19:48 — Tuberculosis-related names and their consequences for clinical interpretation
- 00:21:29 — The E. coli–Shigella example of genomic grouping versus clinical naming
- 00:24:06 — Why not every newly identified clade needs a lineage name
In their own words
I think we’re at a turning point in taxonomic history, and it will change eventually, even if not now.
— Leighton Pritchard, 00:14:58
You cannot force somebody to use a new name.
— Conor Meehan, 00:15:03
Who is talking
- Lee Katz (host)
- Nabil-Fareed Alikhan (host)
- Leighton Pritchard (guest, Strathclyde Institute of Pharmacy and Biomedical Sciences, University of Strathclyde)
- Conor Meehan (guest, University of Bradford)
Also mentioned: Ford Doolittle.
Tools and resources mentioned
pyani, NCBI Taxonomy, 16S sequence comparison, Average nucleotide identity (ANI), DNA–DNA hybridisation (DDH), Digital DNA–DNA hybridisation.
Questions this episode answers
How does the episode define a bacterial species?
It does not offer a single universal definition. The guests discuss discontinuities in gene flow alongside the practical use of biochemical traits, phenotypes and genetic similarity to distinguish groups.
What ANI threshold is discussed for bacterial species?
Meehan gives approximately 95% ANI, relating it to historical thresholds of 97% 16S similarity and 70% DNA–DNA hybridisation. The episode presents these as approximate correspondences, with ANI comparing shared genomic regions.
Can a metagenome-assembled genome be type material?
In the formal system described in this 2021 episode, proposals to accept genome sequences as type material had been rejected by the ICSP. The guests expect the rules to change because culture requirements exclude many organisms.
Why keep Shigella separate from E. coli in clinical naming?
The panel describes Shigella as genomically embedded within E. coli but clinically important to recognise separately. It uses this example to show why useful medical names may not map neatly onto genomic species boundaries.
This page was written from a machine transcript of the episode and checked against it. Listen to the episode for the whole conversation.
Show notes
We’re navigating the twisted world of bacterial taxonomy. We have some excellent guides to help us!
Our guests today are:
-
Dr. Leighton Pritchard: Who is a Strathclyde Chancellor’s Fellow at Strathclyde Institute of Pharmacy and Biomedical Sciences in the University of Strathclyde
-
Dr. Conor Meehan: Dr. Conor Meehan is an assistant professor in molecular microbiology at the University of Bradford.
The conversation continues in part 2.