Talk:Angiosperm Phylogeny Group

I removed the following reference, inserted by an anoymous user. It is not cited in the text and it is not obvious why it is relevant. If you decide to put it back in the page, please explain here why - and mke sure the text tells the reader why it is relevant:

• G.E. Gvaladze, "The Chalazal Polar Nucleus of the Central Cell of Angiosperm Embryo Sac", Publishing House "Metsniereba", Tbilisi, 1976, 138 pp. (a monograph. In Russian, English summary)

seglea 19:00, 26 Aug 2004 (UTC)

Origins, history, and process of the APG

On reading this, I am looking for more on how APG was originally convened, what its organizational stucture is, and what is the process it uses to maintain and revise its work. underalms 15:40, 13 May 2009 (UTC)

I agree. However, I think this is a difficult article to re-write (I agree that at present it's not very good); I have tried but failed to produce something which satisfied me. One problem is the multiplicity of articles in Wikipedia. There are separate articles on the output of the APG (i.e. APG system, APG II system and APG III system). Inevitably there will be overlap. Further, changes and developments aren't clearly described when split up in this way. Thus the article on the APG II system needs to be altered, since it implies that it "is" the system produced by the APG, whereas now it "was" an earlier system. Wouldn't a single larger article on the APG and its work be better? Then there could be sections on e.g. "Origins", "Composition", "Purpose", "APG I system", "APG II system", "APG III system", and so on. Peter (talk) 09:59, 4 November 2009 (UTC)

Angiosperm Paraphylogeny Group

It is just for to be based on chloroplast genes that the main reason which this classification shall be rejected because it analyses the plastid genome (thus an allogenome), a cyanobacteria, a co-evolutional being, a symbiont, and not the plant genome per se (by itself); and regarded chaotic (since it makes changes so radically, that soon, will be better that we use only the common names rather than scientific names)(see also Talk:PhyloCode#PhyloCode and nomenclatural chaos), anti-Linnaean (since it uses Cladistics, that is, does not fit in Linnaean taxonomic categories)(see talk:Cladistics#Cladistics as method not compatible with Linnaean Taxonomy) and anti-Taxonomic (since it is corrosive to major taxon, like Family).Berton 18:11, 2 January 2006 (UTC)

The study of genome of the chloroplast has been disturbed for the fact of this genome it is uniparentally inherited and this on the one hand technically facilitates the analysis on the other hand it changes the phylogeny inference, consisting of paraphylogeny. (see http://www.ndsu.nodak.edu/instruct/mcclean/plsc431/maternal/maternal3.htm "This is conclusive evidence that chloroplast DNA is inherited in a maternal manner."]

Then the studies based on chloroplast DNA are evaluations of the endosymbiont genome that has inheritance pattern different from the plant genome (host).

See also Horizontal gene transfer#Evolutionary theory. Berton 22:25, 7 March 2006 (UTC)

Observe that: Bacteria of the genus Agrobacterium have developed a special mechanism that allows transfer of genes from the bacterium to higher plant chromosomes. Berton 12:25, 8 March 2006 (UTC)

Other allogenomes: DNA transposons, retrotransposons, retrovirus, B chromosomes, plasmids, other endosymbionts; moreover most of DNA is junk DNA.Berton 20:07, 9 March 2006 (UTC)

It is concluded that: after all these confused factors for the genetic research that this no doubt it should be subordinated to the morphological research, that is a field of the solid knowledge and not subject to "jack-in-the-box factors" (surprise factor) as the already mentioned. APG provoked an unprecedented crisis in the history of the Taxonomy when giving priority for the genetic research and relegating to second plan the morphological research. Kubitzki system has been making something very different having a balance between the two research types and it is for that reason that should be considered better. Berton 12:49, 8 March 2006 (UTC)

So does this mean that you will accept the Kubitzki system of classification for Malvales, and Malvaceae in particular? (BTW I need to point out that the APG itself is not provoking any kind of "unprecedented crisis" as it is simply pulling together the research being done by many different researchers at many different institutions, who are all getting much the same results with regard to plant phylogeny, using many different kinds of methodology, and which are generally supported by morphological, anatomical, biogeographical, and other non-DNA characters.) MrDarwin 13:43, 8 March 2006 (UTC)

MrDarwin, I tend to accept the classification of Malvales sensu Kubitzki. As the crisis provoked by APG, it is enough to give a glance in its radical changes that it proposed (with enormous quantity of genera removal) and that will probably continue proposing (nothing guarantees that APG III will not substitute APG II and so on).Berton 14:27, 8 March 2006 (UTC)

Berton, I see you have restored the text "In practice, not all DNA data is created equal. Fortunately the DNA most plentifully present in plants, that of chloroplasts proved to be very informative. Actually this is not plant DNA, but prokaryote DNA." I removed this because it reads like a non sequitur, doesn't add anything informative and seems to be injecting a point of view. Plenty of information on chloroplasts and their DNA can be found in the linked chloroplast article. The nature of chloroplast DNA is irrelevant to this article, since chloroplasts and their DNA have been integral parts of the plant genome for the entire history of plant evolution. If you think the commentary is both relevant and necessary to this article (which is about the APG, not about construction of molecular phylogenies using chloroplast DNA), I would suggest rewording it to better express how and why you think this text is relevant to this article, while keeping a NPOV (for example, can you provide a published reference that criticizes the APG for their use of chloroplast DNA to construct phylogenies?). In other words, any criticisms of the methods and conclusions of the APG should not be your own, but I think published criticisms are fair game. MrDarwin 15:18, 9 January 2006 (UTC)

MrDarwin, I understand your position (you want the best redaction for this article, without provoking controversies with point of views non proven), however, I was not who wrote these phrases, just found important to conserve them because they show where the classification preconized by APG is based (thence your relevance).The cloroplast genome is not a plant genome, therefore it is called allogenome (allo means: from other) although plastids have been integral parts of the plant in the entire history of plant evolution (as endosymbionts), it is the same that to consider the bacteria genome of the human gut flora (also a symbiont) as human genome, that is so obvious that I don't need to mention sources.Berton 11:55, 10 January 2006 (UTC)

It may be "obvious", but it's also wrong. The chloroplast is a cell organelle of endosymbiotic origin, and its genome is passed to offspring in much the same was as plant nuclear DNA (except for being uniparentally inherited), while the human gut flora are independently reproducing organisms. One should recognise that the cpDNA tree may not correctly represent the relationships of nuclear DNA, due to the effects of hybridisation and introgression, but to same that it's the same as the human gut flora is nonsense. [S.R. Hinsley]

I agree (it was just an analogy), but I conserve the essential: cp genome is not plant genome, and this is not phylogeny, but paraphylogeny. Berton 11:22, 25 February 2006 (UTC)

I have rewritten slightly to (I hope) make things a bit clearer. As I said previously, any published criticisms or commentary on the APG II methodology or conclusions are fair game to reference, but unreferenced commentary is inserting a POV. I left in the reference to phylogenies being based on chloroplast DNA (which links to the chloroplast article) but the rest is superfluous. Any comments on or criticisms of the use of chloroplast DNA in constructing phylogenies, evidence of chloroplasts not evolving in perfect parallel with the nuclear genome in plants, and/or references to chloroplast DNA producing incorrect or misleading phylogenies, should go under the chloroplast entry, where more information on chloroplast DNA is needed anyway. Perhaps a separate entry on constructing molecular phylogenies in general would be useful, and that could go into the pros and cons of using various kinds of DNA. MrDarwin 15:02, 14 January 2006 (UTC)

And now Brya has reinserted the text "In practice, not all DNA data is equally useful. Fortunately the DNA most plentifully present in plants, that of chloroplasts proved to be very informative. Actually this is not plant DNA, but prokaryote DNA." This particular article is not about chloroplast DNA, it's about the APG. I would suggest that commentary on, or criticism of, chloroplast DNA and its usefulness in constructing phylogenies belongs in an article on chloroplasts, on chloroplast DNA, or on molecular phylogenies, rather than here. Brya (or Berton), if this information is so important to this article, please SAY WHY. The text as written reads awkwardly and as a non sequitur. So please say WHY (preferably with references) "not all DNA data is equally useful", and please say WHY "this is not plant DNA, but prokaryote DNA" is significant. MrDarwin 15:23, 23 January 2006 (UTC)

It is very significant because the system of classification of APG is based on the research of chloroplast genome, that is not of the plant in itself, and for that it should be rejected.Berton 15:50, 23 January 2006 (UTC)

Berton, that is your POV and does not seem to be shared by the majority of botanists. There is a broad consensus among systematic botanists that chloroplast DNA is suitable (with some caveats) for reconstruction of angiosperm phylogeny, and I think there is also a pretty broad consensus that the phylogenies presented by the APG are getting close to the real phylogenetic relationships of extant angiosperms. Of course translating a phylogeny, however accurate, to a classification system, and applying a nomenclature to that system, are quite another matter--but that doesn't seem to be the part you are objecting to. MrDarwin 18:07, 23 January 2006 (UTC)

MrDarwin, we must to have critical sense, why is not most missed, that I should also, the phylogeny must to reproduce more exactly possible (without any error) the evolution and not the co-evolution and thus it should not be a paraphylogeny instead of phylogeny. I already said that classifications are merely suggestions, but as the APG System is entirely based on allogenome research, a great error and therefore it should be rejected.Berton 18:42, 23 January 2006 (UTC)

Perhaps I haven't made my own position clear. I am not proposing that Wikipedia "accept" or "adopt" the APG II system, but neither do I think it can be "rejected": all of those are POV's and I am disturbed by statements I have seen in various articles that Wikipedia has "adopted" the APG II system. In my opinion, all Wikipedia can do is describe the APG and what the APG was proposing. There are other classification systems in existence and I don't think Wikipedia can or should pick and choose one of them as more valid than the others (although it can acknowledge whether one system or another has been accepted or rejected by consensus of systematic botanists, which point I don't think APG II has reached quite yet, in part because this field is in rapid flux and any system is going to be out of date almost as soon as it is published). But that's all putting that cart in front of the horse: this article is not about the APG II system and whether it should be accepted or rejected; it is about the APG, who they are, what they published, and what they did. But I might note that the reason you are giving for rejecting this system (that it is a "paraphylogeny") is not a reason most systematic botanists would give for rejecting it. That is your opinion, and as such does not belong in the article. MrDarwin 20:08, 23 January 2006 (UTC)

Well, this article is about the APG, and you could argue that it should not include information on the APG system, or how this came about. This would lead to separate entries on both APG I and APG II, with various links. Basically I am not against that, although I am not volunteering.

The information on the genes used, and where these genes occur, is relevant because people tend to buy into the propaganda of "DNA is the code of life and by looking at the DNA we will know all" or something along those lines. The genes used are so important because it makes it clear that the differences in DNA sequences are not genetically expressed. The genes function perfectly and identically regardless of DNA sequence.

I disagree with Berton that APG would lose in value because these are not 'real' plant genes. Basically, chloroplast DNA is just one more source of information, to be evaluated in its own right. As it happens it is a very good source of information (at higher levels), regardless of the style of the APG group in classifying. Brya 21:58, 23 January 2006 (UTC)

Latest revision

For the most part I'm satisfied with Brya's latest changes; I may do some minor cleanup & editing but don't expect to have much time in the next few days so the changes may stand as is. Although Brya is correct that the Cronquist system was not "almost entirely" based on morphology (calling it "myth" is a bit strong, I think), I would contend that Cronquist, following on the heels of virtually every preceding major classification system, relied primarily on morphological (under which heading I include anatomical) characters, with chemical and other characters playing a much smaller role. Cronquist's reliance on morphology (and anatomy), combined with his non-cladistic approach, led him to recognize numerous paraphyletic or polyphyletic groups (e.g., Loganiaceae, Scrophulariaceae, Liliaceae, and virtually every order) but that discussion probably belongs under the Cronquist system article. The important part here is simply that the APG and other recent classifications are using a completely new kind of data unavailable to previous researchers. MrDarwin 17:58, 31 January 2006 (UTC)

I've deleted the following text inserted by Brya:

Sometimes overly popularized propaganda claims that APG uses "genetic evidence" or "genetic research", but this is blatant nonsense. A difference in DNA sequence is only usable to determine evolutionary relationships if this difference is not genetically expressed. Genetics and genomics are quite different fields from plant systematics.

I'm really not quite certain what the point of this is, what it's trying to say, or what its relevance is to this article. Maybe it would make more sense with less hyperbole and re-written to make the point more clear. I would certainly want to see documentation of the claim that "a difference in DNA sequence is only usable to determine evolutionary relationships if this difference is not genetically expressed", a claim I believe is overly broad. Discussion (and critique) of molecular techniques is probably best under the molecular systematics article, which can be referenced here. MrDarwin 23:34, 15 March 2006 (UTC)

Well, this bit has a point, which is that all through Wikipedia there are references to "genetic research has shown that" or "genetic data have proven that" when referring to molecular systematics. After finding such a claim newly inserted in the Tiliaceae entry I tried to discourage this widespread practice by inserting the said paragraph.

I will admit this has only marginal usefulness for the APG lemma itself, as it expounds a negative, and this never is good.

As to "a difference in DNA sequence is only usable to determine evolutionary relationships if this difference is not genetically expressed", a rigorous discussion and documention of this axiom would be going rather far. Brya 08:01, 16 March 2006 (UTC)

I'd prefer to refer to "DNA sequence data" or "molecular data" rather than "genetic research" - it seems more precise. Lavateraguy 01:23, 17 March 2006 (UTC)

Either I'm misunderstanding you, or I am missing something very fundamental about molecular systematics (which I will admit is not my specialty). But there is no need for a rigorous discussion, please just point me to a reference or a website that will provide the information or at least clarify your claim. I have been combing the literature, and haven't been able to verify it. On the contrary, rbcL and ITS sequences are two commonly-used examples in which differences in nucleotide sequences are expressed as differences in amino acid sequences. MrDarwin 17:49, 16 March 2006 (UTC)

Differences in ITS nucleotide sequences are *not* expressed as differences in amino acid sequences. ITS is not translated to protein. It is part of the nuclear ribosomal RNA (nrRNA) array; the transcribed RNA genes (5.8S, 18S, 28S) are incorporated into ribosomes, and the spacers (ITS1, ITS2, ETS, IGS) are presumably disposed of in some fashion.

I wonder if Byra means phenotypically expressed. Anyway, it seems to be wrong. There are several properties which mean a locus is better for taxonomic studies, such as ease of sequencing (which is one reason why cpDNA, mtDNA and nrDNA are popular), and an appropriate level of variability. Neutral changes are better, as traits under selection are vulnerable to convergence, but that doesn't make the data invalid for taxonomic purposes anyone more so than morphological traits that are under selection would be invalid. In at least one way protein coding sequences are better - they are more easily aligned, so there is less risk of errors resulting from mistakes in alignment. S.R. Hinsley, Lavateraguy 01:23, 17 March 2006 (UTC)

I would have thought rbcL is a perfect example. Across the plant realm, there have been I don't know how many different sequences of the gene registered. This results in a fairly limited number of different amino acid sequences. Looking at the working of RuBisCO how many different ways are there that this enzyme functions? That is, how many differences in measurable genetic properties of the plant? All in all, compared to the many differences in DNA-sequences that exist there is practically no effect in the genetics. RuBisCO will do what it always does.

And here we have all those wikipedians equating "APG" = "genetics"! Brya 21:23, 16 March 2006 (UTC)

If a nucleotide difference results in a substitution of one amino acid for another, then that difference is being expressed. Does that amino acid affect the functioning of the protein? Maybe, maybe not (I could dig for an example, but I think I've made my point, and in the meantime you've not exactly bent over backwards to support your own statements even when asked!). But it looks like we're working under two different definitions of the word "expressed". The bottom line is that molecular systematics does not rely solely on silent mutations. MrDarwin 23:43, 16 March 2006 (UTC)

OK, you made your point. Indeed, my phrase "A difference in DNA sequence is only usable to determine evolutionary relationships if this difference is not genetically expressed." is an overstatement. The fact that a difference in sequence is expressed, either in amino acid sequence or in genetic properties does not disqualify it per se from being phylogenetically usable. On the whole there is a gradient of the relationship between DNA sequence and "phylogenetic usefulness", and few absolutes pertain. However, on the whole molecular systematics will deal mostly (but indeed not quite exclusively) on silent mutations. This will be even more true in the case of APG, which uses just the three genes (at least in APG I). Three genes which code for proteins that are extremely well conserved, allowing for only a very low degree of variation (certainly well below what is statitically significant). Which of course is just why these three genes have proved useful for the purposes to which APG has put them. How about the text:

"Sometimes overly popularized propaganda claims that APG uses "genetic evidence" or "genetic research", but this is not correct. A difference in DNA sequence of the genes involved here will have no consequences in the properties of the plant that could be called genetically significant. Genetics and genomics are quite different fields from plant systematics." Brya 10:27, 17 March 2006 (UTC)

I still don't see why this needs to be included at all, maybe because "overly popularized propaganda claims" is pushing all my POV buttons, especially without citation of any such claims. I agree that it's important to explain which genes were used and why (e.g., the fact that they are slowly-evolving, i.e., highly conserved and show relatively little variation, and even less expressed variation, is precisely why they are useful for deep phylogenies; a phylogeny of species within a genus will use genes that show much more variation, including phenotypically expressed differences) but anything beyond that starts to sound like critique rather than description or explanation. As I said to Berton much earlier in this discussion, published critiques of the APG methodology or conclusions are fair game--go for it. But the article should not reflect the editor's opinions. MrDarwin 18:38, 17 March 2006 (UTC)

The reason I feel this needs to be included is the widespread usage throughout wikipedia of "APG = genetics". This needs to be combatted somehow. Probably by this time this misconception is rooted so firmly that it can no longer be eradicated, but it needs to be discouraged from growing further. How about:

"Claims that APG is using "genetic evidence" or "genetic research" are not correct. A difference in DNA sequence as used here has no perceptable consequences in the genetic properties of the plant. Genetics and genomics are quite different fields from plant systematics. The accepted term for "DNA sequence data" is "molecular data"." Brya 22:01, 17 March 2006 (UTC)

I don't see why the claims are incorrect. They are imprecise, and "DNA sequence data" or "molecular data" are more precise alternatives (though "molecular data" could be taken to be referring to the metabolonome, particularly those secondary metabolites produced by particular plants), but isn't DNA sequence data a subset of genetic data? On could argue that for a general audience "genetic" is better. "Genetic (DNA sequence) data", anyone?

BTW, as of 1997, pace sequencing errors, 78% of residues in the RuBisCo LSU were known to vary, fide Kellogg & Juliana, American Journal of Botany 84(3): 413–428. 1997 Lavateraguy 13:34, 18 March 2006 (UTC)

I've not seen any articles equating APG with "genetics" but I don't doubt there are some out there. Of course anybody reading such an article can click on the link to the APG article for more (and hopefully accurate) information. I agree that referring to "molecular phylogenies" or "molecular data" (with appropriate links for further information) is preferable but I think that for the majority of Wikipedia users, differentiating between "genetics" and "genomics" is splitting hairs. I do agree that all articles should contain factual information, and any erroneous or misleading information should be corrected. MrDarwin 14:39, 18 March 2006 (UTC)

Basis for the system

The article currently reads, "The system is based on two chloroplast genes and one gene coding for ribosomes. This selection of genes from cell organelles is significant; zoological taxonomy similarly uses genes of mitochondria. The genome of cell organelles is separated somewhat from the nuclear genome, both chloroplasts and mitochondria having their own DNA, actually prokaryote DNA. The sequence of nucleotides is subject to a different rate of change compared to nuclear DNA." This appears to be refer to APG (2003)'s substantial reliance on the research of Solstis et al, 2000 (Angiosperm phylogeny inferred from 18S rDNA, rbcL, and atpB sequences. Botanical Journal of the Linnean Society 133: 381–461). However this is by no means the only form of analysis that APG (2003) uses - the article refers to 4-, 5-, 7- and 11-gene analyses, and even some morphological analyses, e.g. Doyle & Endress, 2000 (Morphological phylogenetic analysis of basal angiosperms: comparison and combination International Journal of Plant Sciences 161 (6 Suppl.): S121–S153). The form of words we currently have therefore seems misleading. However this seems to be a surprisingly convtroversial article, and it's not my area of expertise, so I am raising the issue here rather than changing the page immediately; what do other editors think?.

I agree that what was written in the article was misleading. In my re-write, I've *currently* left out all reference to this topic. I think something should be said, but it's not clear quite what, since I see the APG article as an overview/summary, which is expanded in the separate articles on each of the versions of the system. Peter coxhead (talk) 13:07, 13 April 2010 (UTC)

BTW, à propos earlier discussion, it is not "urban myth" but the APG (2003) article that refers to, "...the then widely used modern classifications (e.g. Cronquist, 1981; Thorne, 1992; Takhtajan, 1997)" as being "based on selected similarities and differences in morphology rather than cladistic analysis of larger data sets involving DNA sequences or other forms of systematic data" (p. 400). I think we should go back to describing Cronquist as "Principally based on morphology", as user:MrDarwin suggests above; this would help the less specialist reader understand why APG is so different from the traditional systems which still predominate in field and garden guides.

seglea 23:36, 9 May 2006 (UTC)

Rewrite

Following WP:Bold, I have re-written the article. I've tried to use as much of the previous material as possible, but now re-arranged into sections and expanded. The only bit missing is something about the genes/genetics involved. What was there before is not accurate, as APG II and III have increasingly used more evidence, including multi-gene data sets. I'm not sure whether this should be discussed in this article or in the separate articles on APG I, APG II and APG III. Peter coxhead (talk) 10:26, 13 April 2010 (UTC)

Under Members of the APG, your treatment of Fay is inconsistent with your treatment of the Soltises. Lavateraguy (talk) 11:09, 13 April 2010 (UTC)

Whoops, yes, I hadn't recorded Fay as a contributor to APG I and II. Just a mistake. I'll check both tables again... Thanks. Peter coxhead (talk) 13:04, 13 April 2010 (UTC)

Editorial comment

The section added in this edit expresses an opinion which is unsourced, and seems to be an editorial comment. Further, where is the source for the statement that the APG III system is "still limited to being largely based on just three genes"? Unsourced facts and opinions don't belong in the article. I removed the section, but material was put back. Peter coxhead (talk) 22:36, 9 April 2015 (UTC)

Indeed. the APG systems are based on the findings of many other studies, which often use many more than "just three genes". The first reference in this later diff isn't reliable (and is WP:FRINGE). I can't find any other source that even mentions the second publication, which appears to possibly be self-published -- the ISBN mentioned doesn't appear to exist, and no reliable links turn up when you search for "International Seminar on "Multidisciplinary Approaches in Angiosperm Systematics"". The added text revealed a complete lack of understanding of plant systematics in general. As it stands, the text should not be added back. Rkitko ^(talk) 17:32, 11 April 2015 (UTC)

It's like those maps that say "up to date" across the front and hide the publication date where one is unlikely to find it. That first reference is undated, but has a subheading "The big (un-finished) debate". It appears to be discussing APG I, which would be appropriately dismissed as "so 1998", and is thoroughly superseded even by APG II, let alone APG III. Sminthopsis84 (talk) 19:36, 14 April 2015 (UTC)

On the other hand I don't think it is our place to paper over controversies, but to provide information that allows readers to make their own decisions. Therefore I think the posted sources should be part of the bibliography. APG was never supposed to be the last word, and the impending APG IV exemplifies that. It was the best approach given the information available at the time and being the product of a committee involved compromises. It was rejected by a number of scientists from the start, which is hardly surprising since many of them had their own systems. That a homeopath didn't like it was also not surprising. Popperian scientific philosophy measures the strength of a theory by its refutation, so the controversies are informative. --Michael Goodyear (talk) 23:41, 30 August 2015 (UTC)

Incidentally I did find the conference proceedings that are mentioned above, and they appear to be legitimate.--Michael Goodyear (talk) 00:24, 31 August 2015 (UTC)