Life is not a tree, it’s more of a…


OK, I think the tree of life is obsolete. I have been spending a lot of time looking at horizontal gene transfer, reading about it, looking at it in genomes until my eyes water and my brain dessicates, occasionally blogging about it and soon to be publishing about it. Life is not a tree. To what extent it is not a tree it is debatable, but horizontal gene transfer is pervasive, if not rampant, in all kingdoms.

Horizontal Gene Transfer. Credit: Wikimedia Commons

So we need another term to describe the interconnection of species and genes. In a tree structure, each node can have only one parent, whereas each parent can have many offspring. Sibling nodes always have the same parent. As far as genomes are concerned…. hoo boy. Too many HGT events to count. For any given gene or chromosome/plasmid segment, we cannot reliably assume that it was vertically transferred. So if a node in our tree represents a genome, it may have multiple parents. Hence, not a tree anymore. Rather, a Directed Acyclic Graph or DAG.


A Directed Acyclic Graph, (DAG). Credit: wikimedia commons

However, it may very well be that a genome which contributed genetic material may also be contributing genetic material to the genome that gave it material in the first place. That happens in endosymbiotic events, but not only. Can happen, for example, with many species of bacteria or archaea living in proximity. Contribution of genetic material may be reciprocal. So a more precise definition for life would be a directed reciprocal acyclic graph or a DRAG.

But then, you already know that, don’t you?

(Yeah, and I know that reciprocity technically voids the acyclic condition. Shaddup.)



Share and Enjoy:
  • Fark
  • Digg
  • Technorati
  • StumbleUpon
  • Facebook
  • Reddit
  • Twitter
  • FriendFeed
  • PDF
  • email
  • Print
  • Google Bookmarks

4 Responses to “Life is not a tree, it’s more of a…”

  1. widdowquinn says:

    Weird. Our results indicate that the topology is more of a Bidirectional Interchange Transfer Chain Hierarchy.

  2. Iddo says:

    @widdowquinn I guess we can both agree that it has a Bidirectional Utilitarian Many-to-Many Encoded Representation

  3. Ian Holmes says:

    One thing’s for sure… if you don’t look around once in a while, you may miss it (aka the Fully Extended Reciprocally Reticulate Inference Statement)

  4. William says:

    I understand Darwin suggested the term “Coral”, and as I understand it this would make sense.