Wednesday, October 25, 2006

Minimally-specified organisms

(H/T Telic Thoughts)

At CreationEvolutionDesign, there is interaction from the author with current news about the genetically smallest organism yet found. It is a parasitic bacterium with 160,000 DNA bases and 182 protein-coding genes.

A couple of quick thoughts, in the context of the posts here about probability boundaries.

Firstly, this is a parasitic organism, so it is not capable of independent life. The author of the post points out that non-parasitic organisms generally need nearer 1500 genes. So "simple life" - life not requiring existing organisms to exist - is eight to ten times more complex than this organism.

Secondly, we are starting to home in on a figure for the amount of minimal information for life to exist. By comparing sequences of DNA for shared genes, like Cytochrome C, it ought to be possible to establish a maximum level of specification - that is, the shared sequences should give us an upper limit on how specified a DNA sequence has to be for it to function. It would be possible to determine a lower limit by systematically changing a gene and determining whether it continues to express its enzymatic function - if it expresses no function, then it is insufficiently specified to work. Once these upper and lower limits on specification have been established, we can then make a reasonable guess at how specified this organism is, for example. We can then use the mathematical work that Dembski has done (relating to the UPB and "No Free Lunch") to determine whether or not "life" can be considered to be chance or design.

Finally, there is no point in calculating the probability of correctly picking 160,000 DNA bases - it would be wildly above the UPB, and in any case, nobody is saying that all of the 160,000 bases need to be exactly right for the organism to "work". Let's suggest instead that there are 1800 candidate genes (sequences of DNA) (somehow) already floating around in a sympathetic medium, of which a proto-organism needs to (somehow) collect 180. Even this - which begs the questions of where the genes come from, how genes with suitable sequences might come about, how the control mechanisms might appear, why the medium in which this happens is so amenable to the process - still would be an event which was less probable than the UPB - one which couldn't be classed as a "random" event.