To TB--Taxonomical Genomics?

TB, I'm interested in your proposal (in the Coffee House) that we can redefine taxonomy using your genomics. Have you done some preliminary work about it? Taxonomy did help spawn the concepts of evolution, but evolution has not returned the favor. Instead, taxonomy has gotten a lot harder due to evolution. Try making a taxonomic distinction among incipient species populations Maybe it would be easier if there really are natural kinds...How do you define species, genus, etc. genomically? Genome difference? Fedmahn Kassad recently reported that the wolf (Canis lupus)(which taxonomically include the domestic dog) and the coyote (Canis latrans) had 6% genome difference. They're both undisputably members of the same genus Canis. Can we compare them to another pair, human--chimp, with 1% difference? I propose that either we revise the chimp into Homo troglodytes (just as Linnaeus puts it) or we put the coyote into another genus.

quote:

---

Originally posted by Tranquility Base:
^ Yes all very good points and interesting about the coyote/wolf.

​

The issue that comes out of doing molecular level genomics rather than some sort of DNA fingerprinting etc is that there is more to genomic comparison than 'XX %' as I'm sure you're aware.

​

So it is becoming apparent that the thing to do is look for the occurances of protein families. Within a protein family one might still have large variations but yet the genes are still recognizable as being part of the same family - and doing the analogous thing in that organism. Such variations within a protein familiy can be considered to be allelic. So the idea would be to carefully look, in the genomes we have, for what blocks of genes seem to travel together and what sort of genomic differences you get within the currently defined species, genera, families and higher taxa. Some of this has undoubtedly been done but there is still a long way to go and there may not even be that many people talking quite as you and I are.

​

So one would see if there are some sort of clear patterns and then see if redefinitions of species/genera/families etc allow the patterns to improve. What sort of patterns? OK - one might find that at some level animals are related purely allelically - ie they have the 'same' genes (in terms of function) but diffent alleles. At the next level one might see novel 'luxury' genes: non-essential genes that distinguish a horse from a zebra for example but do not disallow hybridization. Then at higher levels one would look for essential sets of genes that appear as blocks. One would try and see what type of changes one gets comparing acroos classes or orders. I'm thinking aloud here but basically although a lot of this would be partailly known 'anecdotally' it has probably not been done systematically due to the newness of the genome sequence determinations.

​

At some point one will have to think about loss vs gain and try and use pseudo-genes to distinguish between these two. I think it would undoubtedly be a good post-grad project/collaboraiton between a genomicist and a taxonomist.

​

On the other hand you have to take this with a few grains of salt. My genomics work so far concerns (i) protein folding mechanisms (not very genomic really) and 3D structure prediction and (ii) the use of protein folds across pathways. I really know almost nothing about taxonomy - it is only what I have picked up subliminally. And yet I am convinced this is a good project and would be confident of supervising it. I am a fan of multidisciplinary science.

​

Here's a question for you. How exactly do modern taxonomists define new genera or families for example? How clear cut is it? What makes the wolf and dog part of the same genus Canus and not just in the same family? It intuitively makes sense but what are the objective criteria?

​

[This message has been edited by Tranquility Base, 09-12-2002]

---

The species for sexual organisms is defined by Mayr's biological species concept, that is, by reproductive isolation. However the isolation test is not always practical. As for my specialty (termites) I haven't known any researchers defining termite species using BSC. I think the same goes to most invertebrates. Vertebrates are better studied taxonomically. But then again the taxonomist itself is a rare species Some authors use an ecological context to define genera. Bernard Wood & Mark Collard used that definition when they banished Homo habilis and H. rudolfensis out of our genus, based on both species' inferred ecological adaptation; habilis and rudolfensis were considered closer to australopiths than to Homo sapiens, the type species of our genus.However, this ecological generic definition is only guesswork. I can understand that modern humans and Neanderthals and Homo erectus originally have similar niches (albeit we have successfully expanding our niche now). However, wolves and coyotes have different niches, and jackals (Canis aureus and Canis mesomelas) differ with them. The same difference can be found among the members of Panthera. The lion Panthera leo is a pack hunter, the tiger Panthera tigris is a lone hunter and the leopard Panthera pardus is an ambusher. The definition of genus, IMO, is a remnant of how Linnaeus ('the big L.') classify them. This should be revised I think.As for families or higher taxa, I am not sure if there is any consensus between taxonomists. I'll take another example from termites. Tim Myles of the Univ of Toronto proposed that the family Termitidae, which is the largest termite family out of 5, be split into six smaller families. I contacted Paul Eggleton of London Nat History Museum, a leading authority of termites, for his opinion of Myles' proposal. Eggleton said that Myles' only reason was because Termitidae was too big, and that is not a good reason (according to him). From those opinions I am quite certain that the higher taxa (from family upwards) do not have any clear definition, such as Mayr's BSC. That is why I am interested in reading your proposal.I too think that this is something above the controversies.