Ontologies of biology

I would like to read more about the dynamics of our categories. How they are revised and changed.

Guilds, niches...

I'd be happy to read anything you have to say about the ontology of behavior.

Within taxonomy, I'm interested in your thoughts about "character" and/or "homology", specially within the context of the renewed discussion regarding them as kinds (e.g., O. Rieppel) vs as individuals (e.g., Grant and Kluge).

Ontologies of developing anatomy are pretty tough stuff to agree on, since the terms change depending on which developmental stage and often, on which organism. There are some efforts going on in Edinburgh, for example, that are worth citing.

Ack! I have too clever a bunch of readers...

Roberto, that is something I want to address, although Olivier's papers always strike me as undirected, or contradicting what he previously has said. At least he's given up on Popper.

Alethea, I don't want to address the Ontology Project sense of ontology. It has a philosophical meaning (the furniture of some domain of discourse) before it has a database meaning, which I find rather trivial and uninteresting, although the mistakes it makes are very interesting indeed. In particular I dislike the Gene Ontology project.

But if there is a sense in which developmental stages are given generic ontological treatments (like phylotype, although that is fairly restricted to verts), do send me some references.

Niche, guild, community, sure thing.

Behaviour! Wow, I hadn't thought of an ontology of behaviour...

This is looking like a major project. I want to plan it out in some detail, and put drafts of sections up here. If it goes well, then it will become a book treatment. If I'm capable of doing that. Keep 'em coming, guys.

I've been dealing with Gene Ontology quite a bit for the last few years and there are a few oddities in it that I quite like:
-Terms like "Unknown Biological Process" (GO:0008150). There is an equivalent taxa in NCBI taxonomy: TAXID: 32644 (unidentified species). Associating these to a gene or species is like saying "I don't know" vs "I didn't look" which would be no association.
-The ballooning of some parts of ontologies depending on how much attention is focused on that area. My wife worked at a fossil type collection for many years and said that "there are two types of scientists: groupers and splitters". The splitters tend to endlessly subdivide categories, the groupers are more prone to batching stuff together. This leads to a weird, homunculus like scaling of the terms is portions of the ontology (http://scienceblogs.com/omnibrain/2007/07/clay_homunculi.php)
GO deals with this by using GO Slims, an attempt to even things out a bit by reducing it to more uniform parent terms.

Somewhat off topic, just remembered a couple of (quasi) biological classification systems of interest:
-Borges' animal classification system
http://www.multicians.org/thvv/borges-animals.html
-The Vienna Classification of Trademarks which has a subsection for animals in trademarks: delightfully bizarre. I parsed it out and visualized it as a tree map diagram a while ago:
http://www.palidwor.com/blog/?p=77
My favorite class is: "Reptiles, Amphibia, Snails, Seals, Sea Lions, all grouped together.

Borges invented his Celestial Encyclopedia. The Vienna thing is a very good representation of the half-morphology, half-habitat taxonomy of beasts from the middle ages. The giveways are the inclusion of birds and bats in one category (flying things), the groupings of fish including whales (crocodiles also used to fall in that classification, but I bet there weren't many heraldic devices using crocs), and the trashcan category of "everything else" including snakes, sea lions and worms. Classic. You can find precursors to this in any late medieval bestiary.

Arrgh! Arrgh! Ontology! If I never hear the term again, it will be too soon.

Well, maybe it's not that bad. I'm working on the Australian Faunal Directory, which means I work with taxonomists. Now, I'm a Java developer. Not only is taxonomy new to me (beyond a high-school "these are the vertebrates" level), but taxonomy deals with many of the things that computer science is interested in - the interface between data and meaning, the use of names.

Consequently - well, you wanna know in how many different senses I use the word 'type' in an average day at the moment? We got records of a certain type that store type data, and they have a type (lectotype, paratype etc). These types are represented by a (Java) enumerated type, which maps to the (anonymous) enumeration type that's part of the XML type that the 'type' element extends.

(Took me ages to work out that species types are called "type" in the sense that they are typical of the species.)

Anyway. Part of the whole ontology for taxonomy thing is here: http://rs.tdwg.org/ontology/voc/ . Follow the links, and you'll discover the glorious "Semantic Web" and the joys of RDF.

It's all good, clean, fun.

An ontology of objects of selection (gene, parts of organisms, individuals, kin, groups, species, higher taxa, ecosystems) would be interesting, especially because it's scope might extend beyond biology (memes? technology? universes?).

An ontology of "living things" or "organisms" might be interesting too, as a lot of things seem to straddle the boundaries of our normal usage of those terms (viruses, auto-catalyzing RNA, endosymbiotic organelles [mitochondria, chloroplasts], human cells cultured in a lab, HeLa cells, devil facial tumor disease, etc).

You know, there are people who do this sort of thing for a living: here and here and here, ie. Biological knowledge representation has been a field of research for decades.

ortholog vs. paralog