We are all one. One Spider, apparently

[andrewducker]

Nice article at Wired News: Linux Reconstructing Tree of Life.

Wheeler's team will work to create a massive data matrix containing detailed information about each spider species -- attributes like size, diet and habitat. The matrix will then be processed to spot subtle relationships between the species, and to figure out how spiders made the transition from their family tree's base to the farthest branches.

Demeter will then analyze the data arrays to arrive at the best tree formation, among many possible trees, that could describe evolutionary relationships.

I'm intrigued as to why they're using non-genetic methods of relating them, especially when there's a move to classify species according to genetic similarities.

Comments

[kpollock.livejournal.com]

I think that they may be concentrating on phenotype and behaviours, as many of these can be the same for differing genotypes. Also a change in genotype is not the complete reaction to the enviroment - the outcome (i.e. behaviour and physical qualities) is what matters in terms of eating/getting eaten/squished/breeding etc. many genetic changes alter nothing that affects these fundamentals.

hastily put and badly, no doubt - I'm finding it hard to express the concept that crystallises in my head.

Genetic similarities can also be co-incidental.

[andrewducker]

You're right, genotypes can backtrack and explore solution domains without causing actual change (Mike's told me some fascinating stuff along those lines), and it would be very interesting to look at links from both points of view and see what differences there are.

[heron61.livejournal.com]

I'm intrigued as to why they're using non-genetic methods of relating them, especially when there's a move to classify species according to genetic similarities.

While useful data has been uncovered using genetics, it still isn't as useful as morphology and behavior for classification. There have been a number of incidents where genetic classification methods have produced nonsense results, in large part because our understanding of genetics is still vastly limited. Standard classification methods are far more robust and should yield very interesting results when applied in this level of detail.

[andrewducker]

It will be interesting, certainly.

I've also read of areas where genetic testing showed that creatures that were thought to be similarly related turned out to be merely externally similar, and vice versa. The medling of both methods of classification should produce some excellent results.

[heron61.livejournal.com]

I was thinking about this and realized that one good reason for not using genetic data was the this same program would be used on dinosaurs, where there will never be genetic data and it would be useful to have a test of the same sort of program. Then, I went back and reread the article and saw that it said we combine information from all possible sources, like anatomy and genomic DNA which pleases me, since using genetic info alone can (and has) produced completely meaningless results, but combining it was behavioral and morphological data could be very interesting indeed.