A place for discussion of the potential adoption of a binomial nomenclature system for virus species names.
Some preliminary thoughts (other comments regarding specific questions in the following thread)Taxonomy is about the organized designation of whatever we are talking about. For viruses, it does make sense to establish a proper manner to designate species names, relying on genera previously established by ICTV, but not necessarily on higher-rank groupings (families, orders etc.). This principle is indeed similar in essence to the use of binomials for other life forms. Before Carl von Linné, species were designated using vernacular names and, in the scientific community, using short diagnostic phrases: quite like viruses today, except that these phrases were not unified for a given species (this unification has been one of ICTV’s achievements in the last decades). This system was almost satisfactory and although Linné did see a urge in listing all godly creatures, there was no consensus about a real need to change their designation. It is merely for the sake of convenience that Linné introduced the binomial nomenclature, first informally as “trivial” names in scientific conversations with his students. Only later did he actually introduced the binomial nomenclature as mere synonymies to more official diagnostic phrases in the first editions of his Species plantarum (Linnaeus, 1753), and finally as full entries in the last ones. The convenience began to spread and lasted until today. By the way, it should be noticed that, in other life forms, there is no contradiction but a complementarity between the use of (Latinized) binomials and of vernacular names. For instance, everyone is OK with “Bird” being the English vernacular name for Aves (and e.g. “Oiseau” the French one, “Vogel” the German one, etc.) while within Aves, “Nightingale” is the English vernacular name for “Luscinia megarhynchos” (and e.g. “Rossignol” the French one, “Ruiseñor” the Spanish one, etc.). In the virological community, there is currently a plain confusion between vernacular and scientific names, even in English (Van Regenmortel et al., 2010): the status quo is not fully satisfactory. Any attempt at designing a uniform manner to name virus species should cross existing Taxonomy, biological features, and a clear prospect for adoption in every community of virologists (phages, plant viruses, animal viruses, human viruses, etc.). Earlier attempts have failed, for a number of reasons related to one or more of these dimensions. Indeed, there has already been attempts to introduce Latinized binomials but these did not necessarily rely on genera and species, or other sensible properties, and they were not adopted. See for instance the result of one of these early attempts, in the title of this article from the 1st half of the 20th century (Best, 1940): “Methods for the preparation of pure tobacco mosaic virus nucleoprotein (Marmor tabaci var. vulgare, Holmes)”. Marmor tabaci (Tobacco mosaic virus) and Marmor cucumis (Cucumber mosaic virus) never belonged to the same genus –nor even to the same family– and in any case no such thing as a genus or a family of plant viruses was identified back then; therefore genus names like Marmor did not make any sense and were soon duly forgotten. More recently, two types of proposals for binomials were made: Latinized ones (Agut, 2002) or Anglicized ones (Van Regenmortel et al., 2010). ICTV (Siddell et al., 2019) proposes three options for the current consultation. These options differ in the so-called “specific epithet”, used to designate the species within the given genus, but all three rely on the already established list of virus genera. This this shared feature does make a lot of sense and I won’t discuss it further. --> Except maybe to suggest that some distinctive “first name” should be given in the case of species not yet assigned to a genus, in a manner more or less inspired of how “Candidatus” is used to designate non-cultured procaryotes whatever their actual genus (Murray and Stackebrandt, 1995): why not “Unassignedvirus” or an even simpler “Virus”?
Option #1: Latinized binomials
Option #1 of Siddell et al. (2019) is similar to an early proposal with Latinized binomials made by Agut (2002).
Although it would rely on established genera, it has not been adopted so far, perhaps because it has not been considered very practical. Indeed it is not.
Indeed, while this proposal is apparently rational, establishing it would imply a huge piece of work by ICTV experts to design and propose 1000s of new names “out of the blue”. I am afraid I was not convinced by the attempt made by Postler et al. (2016) to demonstrate the feasibility of such an endeavor, nor by its result on the bases of the use cases proposed (the family Arenaviridae and the order Mononegavirales).
If this however succeeds, the inevitably poor connection to already established species names would mean another thick layer of discussions in every community to reach a multitude of agreements regarding each of the Latinized epithets.
When saliva is exhausted and these agreements have finally become widely consensual (if they do), another even thicker layer of efforts can be expected for the entire community to recall and use them. In this respect, quoting Postler et al. (2016), I can hardly consider it to be “an additional advantage” that “Latinized species names will appear foreign to most readers”: I would definitely prefer a naming system that would instead appear familiar to everyone in order for it to be adopted! Just the way the binomial system was adopted some centuries ago for other life forms because it was actually simpler than the already established system of species designation. “No, please no!” is likely to be the only consensus the Latinized binomial nomenclature of viruses will ever reach.
To say it otherwise, more “ICTVwise”, I cannot foresee any good adoption prospect for an option that is so bluntly opposed of the ICTV Code (Rule #3.8 “Existing names of taxa shall be retained whenever feasible”), however smart the name proposals would actually be.
--> Inevitably, this must rule out Option #1.
An unproposed option: Anglicized binomialsAlthough it is mentioned but not proposed by Siddell et al. (2019), the more recent proposal with Anglicized rather than Latinized binomials, made by Van Regenmortel et al. (2010), bears the clear advantage of a good alignment with the current Anglicized taxonomy of species established by ICTV. However it also has some drawbacks, and serious ones I believe. In most cases it would simply not be practical, and/or not actually binomial. For instance, HIV1 would not easily become Human immunodeficiency 1 Lentivirus on a daily basis in the medical communities. Even if it would, it would not really be a binomial nomenclature would it? In the case of HIV1 it would actually be a trinomial-plus-one-figure name, while other species would be attributed quadrinomial names (e.g. African horse sickness Orbivirus), or even worse. Hum. In addition –and in fact, just like the current system–, this proposal would lead to the elaboration of complicated expressions in non-English languages (yes, there is a virology literature in vernacular languages!). For instance, in French, the sentence “the plant is infected by CMV” would become “la plante est infectée par le virus de la mosaïque du concombre (Cucumber mosaic Cucumovirus)”. This rather complicated combination of the vernacular and the Anglicized scientific names would inevitably become the doubly incorrect –yet currently quite frequent– “la plante est infectée par le Cucumber mosaic cucumovirus”, where the vernacular name is erased/crashed by the Anglicized scientific name that is used, instead of as a proper noun, as a common name (see the presence of a definite article) however in italics. What a (cucumber) soup. --> For these reasons at least, I would not favor such a proposal either.
Option #2: alphabetical characters in a logical seriesI do not think Option #2 of Siddell et al. (2019) is practical enough, despite its apparent extreme rationality. Namely, I do not see which kind of “logic” would be able to lead a “logical” series of alphanumeric characters. An alphabetical order for the already described species, then the chronological order of the discovery/description of a future virus species, most probably? But then, as the ICTV proposers list as the 2nd disadvantage, how could subsequent taxonomical reorganizations be made, when necessary, without disordering such series? --> No: Option #2 has to be ruled out as well because it lacks flexibility. It is too rational, in a way.
Option #3: freeform textHaving disliked the status quo and excluded Options #1 and #2 of Siddell et al. (2019) as well as the unproposed option of Anglicized not-really-binomials of Van Regenmortel et al. (2010), what else remains? Let us consider Option #3. Although I first thought it was a strange proposal, I may be able to imagine some forms of Option #3, “freeform text”, that could probably be workable. There is currently an existing and already adopted alphanumeric string attached unequivocally to each virus species: its acronym (or if there is no acronym, one can probably be created from full the species name in an easy and non-controversial manner). Each community of virologists and/or stakeholders is familiar with the use of these acronyms, which are sometimes even more easily remembered than the full species names (scientific as well as vernacular ones). Therefore, using the current acronym as an epithet could be an easy, efficient and straightforward way to proceed: Lentivirus HIV1, Cucumovirus CMV, Potyvirus LMV, Mammarenavirus Lassa, etc. This would:* Require a relatively limited effort for ICTV to establish the list of epithets, because most often the acronyms are already in use;* Not trigger lengthy discussions about the choices of one epithet vs. another, because most often these choices have already been made;* Be as pronounceable as the current acronyms, the broad use of which is an indication that they are practicable (even if, I admit, usually not very elegant);* Be adopted quite rapidly by all communities, since they have actually already been adopted. Furthermore, coming back to the above example in a vernacular language (e.g. French), I could even imagine how sentences like “La plante est infectée par le virus de la mosaïque du concombre (Cucumovirus CMV)” could promote the proper difference between the vernacular and the scientific names, and therefore improve the current practice of naming virus species in every language and every community, not only academic ones. Note that better distinguishing between the vernacular and the scientific names, while it is an issue probably more easily perceived by a non-native English speaker, would actually also benefit the English speaker (Van Regenmortel et al., 2010).--> Option #3 might be workable.
To summarize1. If a binomial system were to be finally proposed for adoption I would therefore suggest it to be in the form: italicized [Genus + Acronym] for already established species, and [Genus + whatever but in most cases the acronym] for species yet to be described. 2. For species not yet assigned to a genus, “Unassignedvirus” or more simply “Virus” could be used as the first term (italicized to distinguish it from the object “a virus”).It is of course uneasy to imagine how such a practice would turn out in a few years’ time, when new virus species are created for bona-fide viruses as well as, perhaps, for OTUs (the biological significance of which is uncertain, and will probably remain so for some time), or for any other type of entities future virus taxonomists would like to name. It is however easy to imagine how such a “freeform” system would bear enough flexibility to accommodate all foreseen as well as unforeseen evolutions of virus taxonomy.
ah, some sense at last from the great Olivier
Triggered by your initiative on Twitter, Prof ;o)
I'm doing my very best to stimulate the discussion, in my own little way . ;-)