4 ms·
I honestly thought this article would be about the other kind of trees - the "trees of life". Because these aren't trees either. General population is just taug
by TeMPOraL 6y ago
I honestly thought this article would be about the other kind of trees - the "trees of life". Because these aren't trees either. General population is just taught they are, which is one of those categorization that doesn't correspond well with the world.
Life is just one big chemical reaction that's been going on for a billion years. It's structured only so-so - not perfectly. In sexually reproducing organisms, genes of two parents mix up randomly(ish) to create offspring, which would suggest a tree structure - except when it doesn't. Sometimes genes mix wrong. Gametes can and do accumulate random mutation. Viruses occasionally inject genes into places where they can get carried to future generations. Incest happens. Not all life forms reproduce sexually. Cells can pick up free-floating genetic information from the environment. At the level of single-cellular organisms, this is arguably the bulk of what's happening. Bacteria exchange genes like people exchange cooking recipes.
The way you take this messy reality and arrive at a "tree of life" (and it's always a tree of life, not the tree of life) is by looking at how some set of genes seem to have spread across life forms. It's always conditioned on what particular genetic marker you're looking at, and is equivalent to selecting an interesting spanning tree in a highly interconnected graph. Pick a different genetic marker, you'll get a different tree.
Sure, once you get to studying trees and dogs and humans, gene transfer becomes more tree-ish - but it's a difference of degree, not of a kind.
- owl57 6y agoIn sexually reproducing organisms, genes of two parents mix up randomly(ish) to create offspring, which would suggest a tree structure. No? Tree structure is a simplified model of asexual reproduction, where each organism has one parent, not two.
- wongarsu 6y agoAgreed. Sexual reproduction suggests a unidirectional acyclic graph. Unidirectional acyclic graphs also allow for Neanderthals reproducing with humans, for random mutations, for bacteria exchanging genes asexually, and for genes inherited from virus infections. But "tree of life" is just so much more sexy than "unidirectional acyclic graph of life"
- owl57 6y agoDAG of life sounds pretty sexy if you ask me, but yes, I wouldn't try selling that anywhere outside CS or bioinformatics department.
- xipho 6y agoWhy not? I've heard it before in conversations. Search "inferring species networks from gene trees". Look at the math in RaXML, MrBayes, etc. The intersection of two different fields is often the most interesting, if you can communicate between both.
- masklinn 6y agoAnd at the phylogenetic level even a dag might be too restrictive due to horizontal gene transfers.