Sunday, September 1, 2013

Talk of the day

Definitely goes to one Soren Vedal. And not just because he gave one of the two talks that came in under time - seriously, is it me or a cultural thing or something else entirely that makes me hate it when speakers go over time? It's not that they haven't know for several months how long their talk should be and haven't had time to practice. It wasn't terribly problematic today, but still.

Anyway. The talk was in the session about temporal phenomena on a biological time scale. It communicated a system level concept, cleanly and efficiently in an engaging manner.


Suppose you had a population of bacteria. At the population level (definitely not the individual level) it makes sense to sacrifice some of your, less fit populace, so that there are more resources for the better adapted to survive. This being a basic trade off called hedging your bets.

This can be improved upon though. When bacteria divide, they don't divide evenly. A disproportionate amount of any damage that a bacteria may have suffered is shunted off into one of the two offspring. This results in one bacteria in better shape and more able to reproduce than another. Which in the longer term means taking less time to replicate and the proportion of undamaged cells increases. More environmental pressure exacerbates this effect. It is basically diverting you damage that a population suffers off to into a redundant evolutionary side track and out of the population. The average damage goes down and the population growth rate also slowly increases. Which is better than static bet hedging.

And all backed up with experiments that fit the predicted patterns.

Nice theory. Nice system. Well communicated.
 

2 comments:

  1. You wrote :
    "When bacteria divide, they don't divide evenly. A disproportionate amount of any damage that a bacteria may have suffered is shunted off into one of the two offspring."

    How does the bacteria "know" which is the damaged bit ?

    ReplyDelete
  2. I wouldn't think there would need to be a mechanism to decide that. It's not that one of the child cells has no damage and the other has all of it. It's more likely I would think that one of the children is less damaged and able to replicate more.

    Very rarely I imagine the damage would be distributed evenly in the parent cell, which would lead to the children getting an even split of the damage, but I don't thinks that would mean much on a population level.

    ReplyDelete