Talk:DNA barcoding

From Wikipedia, the free encyclopedia

	MCB Portal This article is within the scope of the Molecular and Cellular Biology WikiProject. To participate, visit the WikiProject for more information. The WikiProject's current monthly collaboration is focused on improving Restriction enzyme.
Article Grading: The article has not been rated for quality and/or importance yet. Please rate the article and then leave comments here to explain the ratings and/or to identify the strengths and weaknesses of the article. Click here for an overview of how to rate this article.

[edit] Balance

I've done a major rewrite, let me know if it sucks. I tried to give some examples that highlight some of the possibilities and problems associated with DNA barcoding. I'm not sure if my referencing is right though as I've never really written a Wikipedia article before. I think the Criticisms section still needs some serious work, more references are needed. Mdhowe 04:38, 3 January 2006 (UTC)

Good rewrite. Putting barcoding in context ("...not a new concept...") is good, as is the bold cutting of some stuff previously there that I had left because I didn't want to step on the toes of prior contributors. The references are also good. I'll make some other changes in the next few days when I have time, and hopefully together we can make topnotch page. Arga Warga 01:28, 4 January 2006 (UTC)

Thanks, I was hoping to get a bit of help. I think this is a really interesting topic and I've been looking for an article to contribute heavily too, it would be nice to get it too featured article quality. What do you think about the 'Case Studies' subsection? I wasn't really sure if it was proper Wikipedia style, as this was adapted from an essay I wrote for uni. Mdhowe 08:09, 4 January 2006 (UTC)

For the record, the "not a new concept" line is misleading. DNA barcoding is explictly not based on determining evolutionary relationships; that is, DNA barcoding is not phylogenetics or systematics (although this is a common misconception). The Woese introduction is therefore out of place. The idea of using a standardized gene fragment to identify all animal species (but not to classify them) is a new idea, although you might make the argument that using genetic data to identify "species" is normal practice in bacteria and archaea.

Would it be incorrect to say that DNA barcoding is a tool used in phylogenetics? --Mdhowe 08:14, 22 February 2006 (UTC)

Yes, it would be incorrect. Phylogenetic methods (NJ trees) are used in barcode analysis, but barcoding itself is very clearly stated (by actual barcode researchers) to not be phylogenetics.

Not to confuse methodology with a scientific discipline... ;-) No, of course can't barcoding be phylogenetics, just as cladistics despite being the unchallenged leader in phylogenetics methodology is not phylogenetics. The tool collects (barcoding) or analyzes (cladistics) the data, the discipline interprets it and puts its hypotheses to the test against it. Still, at present, molecular barcoding (which is a tool suitable for phylogenetics just like an apple-picker is a tool for picking arboreal Rosaceae fruit) is not really used much in phylogenetics (which is a scientific discipline) because researchers are somewhat hesistant to make much of the method until it has matured (which of course "the guy in the lab across the hall" should do, thank you very much). There is a rather harsh critique of the A. fulgerator study around, I think in the Zool.J.Linn.Soc. something like one year ago, which would benefit the article if it were discussed. The A. fulgerator case is especially damning since IIRC scientific essentials such as reproducibility were somewhat lacking, according to the critique, and many of the barcoded "species" were apparently just morphs. Have it around somewhere on my HD, but as butterflies are not my #1 topic (bird-food, that's what they are), someone else could try to dig it up; it might be many months until I happen across it again.

To go back to the original question: Currently, yes. In 10 years' time, maybe not anymore. Dysmorodrepanis 09:39, 11 November 2006 (UTC)

Moved to DNA barcoding instead of DNA barcode, as this will be a more common search string. Evolver 18:34, 4 September 2005 (UTC)

Maybe this was the right choice, but you did it in the wrong way. You should have used the "move" button. Now you have destroyed the edit history for this page. You might want to ask an admin to fix this. Onco_p53 20:44, 4 September 2005 (UTC)

I tried. Wasn't allowed to use move feature. There was essentially no history anyway (only minor edits here and there), and it has not been destroyed at all. If you want it, go here: http://en.wikipedia.org/w/index.php?title=DNA_barcode&action=history. Evolver 20:50, 4 September 2005 (UTC)

It's 'destroyed' in the sense that it's no longer associated with the place where everyone will do the editing. If you can't move a page (because the target already exists) you should ask an admin to move it at Wikipedia:Requested moves. --Saforrest 20:41, 28 February 2006 (UTC)

[edit] need for more citations for statements made in the text

The "criticisms" section, in particular, reads like a personal statement and justification on the part of the editor responsible, as it makes claims and gives figures none of which are backed up by citations - naming anonymous "critics" and then rebutting them looks very much like a "straw man" approach, and is therefore not in keeping with WP:NPOV. Given the large number of critics and the variety of their criticisms, this article would be far more balanced in its presentation if this section were expanded and referenced to a much greater degree. This is not my area of expertise, but I am familiar enough with it (and the debate surrounding it) to recognize that the article is presently giving short shrift to the criticisms and shortcomings of the technique and its applications. Right now, it reads more like a sales pitch. Dyanega 23:23, 29 May 2007 (UTC)

Here is one that I came across in travels from Bird->Bird nest->Protocalliphora. T.L. Whitworth, R.D. Dawson, H. Magalon, E. Baudry 2007. DNA barcoding cannot reliably identify species of the blowfly genus Protocalliphora (Diptera: Calliphoridae). Proceedings of the Royal Society B: Biological Sciences 274(1619):1731-1739 [1] Shyamal 08:21, 27 September 2007 (UTC)

Another reference - Will, K. W., B. D. Mishler & Q. D. Wheeler 2005. The Perils of DNA Barcoding and the Need for Integrative Taxonomy Syst. Biol. 54(5):844–851 [2] Shyamal 06:00, 28 September 2007 (UTC) + S.G. Newmaster, A.J. Fazekas, and S. Ragupathy 2006. DNA barcoding in land plants: evaluation of rbcL in a multigene tiered approach. Can. J. Bol. 84: 335-341 [3]

I have attempted finding the references. The "96% of eukaryotes" seems to be a mis-statement of this below. Shyamal 11:09, 28 September 2007 (UTC)

“

In these initial analyses, 96% of test species were correctly assigned at the phylum level and 100% of insects were placed correctly at the ordinal level (Hebert et al. 2003a).

”

If that's the actual quote, it is actually irrelevant to the section in which it is placed, which refers explicitly to the ability of barcoding to identify species-level taxa. It's simpest to remove it entirely, then. Dyanega 17:56, 28 September 2007 (UTC)

The "Identification of birds" section is especially in need of balance. If you have read the original paper, you'll note that the standard is extremely low. Specimens cannot be traced, location is not taken into account. It is obvious that a Holarctic migratory bird is liable to have more divergent COI barcodes than a restricted-range endemic, while each can be one and only one good species following most all species concepts - namely in a widely-distributed species there may be a high barcoding sequence divergencs AND the population may be contziguous and coherent and reproductively isolated still, with gene flow between subpopulations abundant but on a global level very slow due to the wirde range; the Osprey is a case that comes to mind. Hybridization has not been taken into account.

Basically they apply one of the most uncritical positivist interpretations I've ever seen in a peer-reviewed paper; it actually makes me wonder how this wretched thing passed peer review. Dyanega, you remember the clash we had about the honeybees - you seem rather reserved about these things and I respect your profdessional opinion[*], so can you take a look at the paper (ref 3)? There has not been a published criticism, but the criticism is indeed very fundamental here; I would go as far as to say it barely falls short of glorified bunk science. So rather than waiting til someone has sorted through the large amount of data (which may be never; I get the impression that most avian taxonomists simply don't consider this specific work it worthy of a critique and rather hope it will die quietly) a more fundamental critique of failing good scientific practice would be the way to go. But what to cite here?

[*] Though I still guess you misunderstood me in the honeybee thing - I would not propose the mol-phyl paper to be a study which revised taxonomy, but yet another study which provided yet more evidence which taxonomic approach to prefer. The ecological, behavioral and now also the molecular data support species status for the Himalayan giant honeybees; though they may be morphologically similar to the South Asian ones there seems sufficient evidence of long-standing reproductive isolation - they almost certainly don't mate to any significant degree, making them not just phylospecies but biospecies. I.e. the studies support the taxonomic treatment as proposed by the original describer. Nothing more, nothing less; had species status never been proposed I would never have suggested to adopt it (a remark to the effect "it may well be a distinct species" is in order in such a case, but I try to be careful not to pre-empt proper taxonomic work - though you may not have noticed, since your specialty field is not CJO Harrison's London Clay fossil birds ;-) ) Dysmorodrepanis (talk) 10:59, 5 April 2008 (UTC)

Another thing that entirely goes under in the whole brouhaha - COI barcoding is nothing revolutionary. The sequence is new, but the use of short DNS sequences for taxonomical purposes was extensive in birds in the mid-1990s (though usually for phylogeny in general rather than for species identification). Nobody wants to repreat this these days were it not for barcoding: it has turned out that e.g. 300 bases of cytB are insufficient and too much affected by hybridization to yield reliable data. I have put up a few papers in Testudo (genus) that show how molecular phylogenetics on the species level is done today; multiple sequences and longer sequences and mtDNA as well as nDNA sequences combined with GISing the voucher specimens and optimally also morphometric analyses are considered the "gold standard".

To use a very short stretch of DNA - even considering COI evolves extremely rapidly in parts - may seem expedient, but take away the flashy name and 5 years ago almost anyone in the field would have considered the technique "yesterday's science" and bad science at that. I have seen a lot of quality papers in molecular phylogenetics on the species level, and barcoding papers are not usually among them. And the expanded barcoding protocols are simply molecular phylogeography without the "-geography"; the technique offers (or seems to offer) only a single advantage in keeping the sequencing effort down by smart sequence choice. But with that, it introduces errors.

Especially the belief of some barcoding advocates that phylogeography and thorough analysis of within-population variation is unnecessary is almost farcical. Too many of the supposed "two bird species in one" are based on a sample size of n=2, with no way to tell whether these two specimens were from the same place and time (which would indeed suggest they are two species) or from opposite ends of a large contigous range (which would almost certainly mean that they are not two species but a phylogeographic artefact).

But what to cite? A few of the refs are good in this respect, but most recent critics tend to get involved in mudfighting; to me it seems as if the real issue with barcoding is that as originally proposed it is little more than a rehash of a discarded approach.

It may be that older work like doi:10.1007/s002390010197, doi:10.1007/PL00006481 and doi:10.1016/0169-5347(94)90028-0 could deliver the background needed here (I have not read them). In general, to provide a "deep" critique of barcoding as originally proposed it seems that one has to go back to the mid-late 1990s, when the shortcomings of the "barcoding without the name" approach that was hot in the early-mid 1990s became apparent. Basically, "it must be taken into account that people tried something very similar 10 years before and largely discarded it because the results were too unreliable". I get the impression that the "barcoding battle" is very much a fight about funds and too little a fight about the philosophy of science. But then, I am firmly with Popper on the latter issue and apparently Hébert et al. aren't. Dysmorodrepanis (talk) 10:59, 5 April 2008 (UTC)

doi:10.1071/MF07050 seems another (recent) study that tackles the issue right at the root:

We thus consider that the application of COI barcodes in delphinid cetaceans should be done with caution; not only has the cytochrome b gene been shown to be phylogenetically more informative, but relying only on mitochondrial DNA genes alone may be problematic.

In other words: useful to get additional data, but neither a panacea nor even really new.

It is astounding how few recent papers in the barcoding debate care to discuss the debate of 10 years ago! Dysmorodrepanis (talk) 11:15, 5 April 2008 (UTC)

Also interesting (from 1995): this

An accurately resolved gene tree may not be congruent with the species tree because of lineage sorting of ancestral polymorphisms. DNA sequences from the mitochondrially encoded genes (mtDNA) are attractive sources of characters for estimating the phylogenies of recently evolved taxa because mtDNA evolves rapidly, but its utility is limited because the mitochondrial genes are inherited as a single linkage group (haplotype) and provide only one independent estimate of the species tree.

This is precisely the crux in the barcoding approach - that you'd want a (short and rapidly evolving) sequence (to make it a superior technique from the view of the field researcher with limited time and ressources) but that this will inevitably introduce a particular kind of error - an error you cannot (with the limited data) recognize. I have also seen a nice study on the Bonefish, showing that this may be anything between 2 (or 3) and 8+ species of Albula, but that if a barcoding approach had been followed the result would have been flawed: the correct evolutionary relationships only resolve with heavy-duty sequencing and analytical effort. I can't find this paper again right now, but eventually I'll do and it'll be a reference at Albula (fish) where it can be picked up. Dysmorodrepanis (talk) 07:19, 9 April 2008 (UTC)

This is a reply to the above (Moore 1995) paper, saying it's even worse in particular lineages. One may well check the examples where standard (mtDNA COI) barcoding doesn't cut, whether they fit this bill: "clades exhibiting female philopatry and male dispersal, or a polygynous mating system". Dysmorodrepanis (talk) 07:23, 9 April 2008 (UTC)

Got the bonefish paper - it's here. If you check out figure 2, you'll note that the least one would have to accept as species is Albula vulpes, A. neoguinaica, and (tentatively) lineage "E". Because the latter cannot be unequivocally placed but appears basal to the A. vulpes clade (including lineages "A" and "C"). But to get to that point, it was necessary to run 4 analyses on "564 bp of mitochondrial DNA (mtDNA) cytochrome b from 174 individuals collected at 26 locations". Whereas e.g. the claim of Hébert et al. that Tringa solitaria is more than one species is based on a "648-bp region of the mitochondrial gene, cytochrome c oxidase I (COI)" with a single (NJ, which is the worst possible method in cladistics as it's essentially sexed-up phenetics) analysis of just 12 specimens from unspecified locations of this taxon from a group that is known to hybridize vigorously. :( Dysmorodrepanis (talk) 07:41, 9 April 2008 (UTC)

The problem here is not that your criticisms are unwarranted - far from it - but that it's very difficult to incorporate these references into the article except in the context of "original research", which is not permitted. Without published criticisms of barcoding that use these same arguments and examples, there really is no way to introduce this information - unless these references criticize the use of mtDNA for reconstructing species trees. Since that is what barcoding is being used for, that WOULD merit inclusion in this article, even if barcoding per se was not mentioned. Dyanega (talk) 17:31, 9 April 2008 (UTC)

You'd be correct if I had discussed tis "as is" in the article, but that's why I did it here. A problem with scientific standards exists it seems, somewhat underrepresented in the current debate (which is more about whether essources are drained from other fields of taxonomy etc), but echoing debates of 10 years ago. And if I am right, this would seem to be article material. BUT: it cannot be article material unless there is citeable sources supporting the conclusions.

So people might want to like to go citation-hunting, and I did all this lengthy write-up to make them aware of what to look for. Sundry results of a Google Scholar search for molecular "proper vouchering" can probably be contrasted by the complete lack thereof in the bird barcoding studies. Studies of within-population variation, hybridization, etc in specific taxa can be contrasted with the conclusions drawn for the selfsame taxa to show that conclusions are premature. For example there is massive COI paraphyly in a calidrid in the follow-up study to the 2004 one... Hybridization in shorebirds anyone? (The Handbook of Avian Hybrids of the World should be able to deliver sources, but I do not have this).

As regards the most controversial issue - and I agree that this is thorny indeed -, namely "quick-and-dirty" COI barcoding being a repeat of outdated "quick-and-dirty" cytB sequence snippet cladistics but with the added problem that some barcoding studies (those that don't use apomorphy vs plesiomorphy analysis but only compare seqence similarity outright) are bordering on phenetics: as I said, there are almost certainly citeable sources, but we'd have to dig deep.

For I know that essentially the same thing that is being toutes as brand new was done 10 years ago with a different gene, and it is not being done anymore these days as it was then, and this is because of the methodological drawbacks.

And I'd guess that there was much discussions about that, "molecular taxonomy" was a controversial issue then.

So in conclusion, there should be good sources to point out that a problem exists, even though nobody at present cares to write about it. Not to say in the article "it's all rubbish", but to say "hey look at that, it seems that we have a knowledge/awareness gap in the current research, so caveat lector". SORA might be a place where literature can be found; most of the journals where the debate is now being held didn't even exist back then.

In any case, see here. Highly recommmended. Note the extensive analysis of within-population variation and the proper vouchering and locality information.

In general, I think Talk pages should be used more often to collect studies and other sources addressing actual or potential issues of articles. This approach was extremely successful when we overhauled passerine systematics - the references listed were collected over the course of many months, evaluated, discussed, until they could finally be incorporated in the article in one stroke. The Polymorphism article benefitted from a similar approach. Dysmorodrepanis (talk) 12:26, 17 May 2008 (UTC)

[edit] Tag removed

The tag for the criticism section - suggesting that the section should be merged into the main text did not seem to fit in with the aim of balancing the pros and cons of the technique, something that the separation. The current treatment in the article is definitely better than it used to be - but suggestions could go here. Shyamal (talk) 02:57, 18 February 2008 (UTC)