Why molecular taxonomy is pure bullshit....

Advertise with us
shop deals on amazon
Shop deals on ebay
Slippertalk Orchid Forum

Help Support Slippertalk Orchid Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
Roth

Roth

Well-Known Member
Joined
Jun 10, 2006
Messages
1,469
Reaction score
390
If you look here:


http://images.wikia.com/orchids/en/images/b/b8/Genetic_relationship_and_differentiation_of_Paphiopedilum_and_Phragmepedium_based_on_RAPD_analysis.pdf

Some molecular taxonomists are happy that their 'RAPD analysis results' were in agreement with the morphology of the studied plants. Joyce Hasegawa close to callosum, very far from emersonii and malipoense, besseae hybrids further than besseae is from pearcei... paph rothschildianum close to phrag sargentianum...

What I really LOVE is paphiopedilum emersonii, malipoense and Hamana Wave ( a maudiae hybrid) 'show similar leaf morphology. Really ?

The original Paphiopedilum taxonomy using molecular technics by Cox and Chase looked very similar. During the proof reading, a paph grower had to teach them that it was absolutely crazy, so they reworked their copy...
 
Not all scientists are good scientists as there are bad and good judges and so on. That said it's so stupid it brings a smile to my face.

Their methodology is obviously wrong, perhaps like mixing hybrids and species in their study. Hybrids have already had their genes modified and the machines might interpret that as relations that don't really exist. Something like mixing apples and oranges at the molecular level. Or as we used to say in the early days of computing: ''garbage in; garbage out''

Michel
 
Last edited:
Hmm. This isn't really a systematics/taxonomy paper. Most systematists these days rely on either parsimony or liklihood analyses to elucidate interrelationships. They use a cluster analysis which is one of many multivariate statistics (UPGMA is unweighted pair group method, arithmetic mean- just one version of a cluster). Yes, it does tell us something about similarities, but not the best choice from a taxonomy standpoint. And it's heavily laden with assumptions that the user can dictate. I probably could've used the same data matrix and got a more reasonable dendrogram ('similarity-gram', versus a cladogram= (hopefully) a 'relationship-gram').

As far as the example of the leaves being similar between some barbata hybrid and a Parvi hybrid, it's all in how you CODE the character!!! I'd love to see their coding scheme and data matrix!!! If they simply coded leaves as maybe two or three characters, something like...
*leaves mottled above (yes, no)
*leaves burgundy underneath (yes, no)
YES, these two will show up as being similar, Duh!!! If they were intelligent and added characters for leaf surface structure, these would've separated nicely.

If they wanted to know a TON about leaf morphology, they could've use TPS (thin plate spline) decomposition, now that's CRAZY morphology analysis.

Also, RAPD was shown to work for other genera, and they mention Goodyera. It is well known that different molecules are good/bad for different genera/groups/taxonomic levels. The fact that it successfully separated Paphs from Phrags, but not anything below the generic level tells us clearly that RAPD is NOT a good molecule to discern slippers at the species/hybrid/cultivar level. DUH! I'd say the species of slippers are therefore evolving at a different rate than Goodyera and the comparison is null!!! Double Duh! I wouldn't have published this paper until I found a more logical molecule, this is a "crappy background knowledge paper" that leads to something hopefully better???
 
Heather said:
Interesting that so many names are spieled incorrectly, as well. I realize that is a minor issue comparatively but clearly not a carefully written article.
Click to expand...

Yes!!! That drove me crazy!!! Any taxonomist/systematist worth their salt is so crazy insane over such details. It's like hearing fingernails on a chalkboard to read such errors!!!
 
I'm not an expert but I briefly scanned through the file and it is a big bunch of crap if you ask me. And don't get me started about the "names".

Looking forward to see Guido's contribution to this topic.
 
Read it a little more closely. They didn't actually use morphology in the data matrix, just made inferences back to it. Still, character coding will always be a major point of discussion. Even pretty clear cut things like AGCT for DNA (an example of a multistate character since there are more than two choices)- seems like there are four choices, right? Well, should one weight them based on the probability of one occurring next to another? Richness of one or more in a genome?

Also found it odd in that they said material of Paph delenatii (a parent of Joyce Hasegawa) was not available. Say what? They couldn't get plant material of one of the most commonly grown Paph species???

Authors like these give decent systematists a bad rap!

I'd have to guess that their granting agency (an ag research center if Korea) demanded they publish one or more papers to prove they used the money (wisely???). This is fairly common. If I were that agency, I'd ask for my money back! Curious how many "real" systematics journals declined it before it got to this horticultural journal?
 
Forget what the article says, how do I get some of the phrags!? :D
Taxonomy by DNA information alone is useless, Any knowledgable slipper grower working with this data could have helped them make some sense.
 
I feel sorry for the Ag Research Center that wasted their grant money. I also feel sorry for the editors of a this crappy journal which were desperate enough to print this. I feel ashamed for the referees who let ths thru for publishing. And I hope the authors are still in a pretty early state of their learning curve.
 
John Boy said:
This is truly great scientific reasoning! Congratulations!!!
Click to expand...
I have scientifically rebuked the Cox et all article in several publications, and in the mean time specialists in statistics have shown that the method simply is wrong. Also Cox et al just leave characteristics away that does not fit in their "argument". It is a fact that "molecular taxonomy" (which has nothing to do with taxonomy because it is a purely mathematical technique" gives irratical species level and any level thereunder. And as I have said before ... when I need a lab to differentiate between a Cattley and a Phalaenopsis, or between a blonde and a redhead ... I will quit doing biology.
 
labskaus said:
I feel sorry for the Ag Research Center that wasted their grant money. I also feel sorry for the editors of a this crappy journal which were desperate enough to print this. I feel ashamed for the referees who let ths thru for publishing. And I hope the authors are still in a pretty early state of their learning curve.
Click to expand...
The problem is that all those "molecular freaks" got to be professors in the early nineties and got zillions of dollars to do reserach (also from the AOS) ... they are still getting their big salaries by writing the nonsense ... see Chase, Pridgeon and their croonies,
 
Braem said:
when I need a lab to differentiate between a Cattley and a Phalaenopsis, or between a blonde and a redhead...
Click to expand...

Um, sign me up as a CoI for the blonde/redhead study! I think we should throw in brunettes just to be complete. I think this need serious scientific consideration. And we could do a cluster f@#$ analysis just to make it publishable.
 
Ernie said:
Um, sign me up as a CoI for the blonde/redhead study! I think we should throw in brunettes just to be complete. I think this need serious scientific consideration. And we could do a cluster f@#$ analysis just to make it publishable.
Click to expand...
Definitely, you are on .... and we must differentiate between natural blondes and chemistry blonds ... but again, we need no molecular analysis for that, we just got to seduce them and have a look ..... Maybe there is some grant money about for that ???
 

Latest posts

Back
Top