Re: We believe in design

Iain Strachan wrote:

> I'll probably bow out now as your knowledge of biology is far superior
> to mine, so there's not much point arguing from ignorance ...

I doubt that my knowledge of biology is that superior. I have read quite
a bit on the topic of scale free networks, evolvability etc.

>
> However one comment:
>
>
> On the contrary, fitness surfaces are strongly
> affected by neutrality.
>
>
> Look at for instance RNA and the work done in this
> area. Combine this with the scale free nature of RNA,
> realize that a simple process of duplication and
> preferential attachment can explain scale free
> networks and one comes to understand why RNA sequence
> space extends through and that the common RNA
> structures are close to any other common structure
> within space, connected by neutral pathways which
> extend throughout space.
>
>
> I'll just say that this was not how I understood the two examples in
> the Toussaint paper. I agreed with you that duplication could indeed
> change the fitness surface because it actually alters the dimension of
> sequence space: what is a local minimum in N dimensional space may
> become a saddle-point in (N+n)-dimensional space.
>
Neutrality has the same effect although perhaps not that self evident.

> But the two examples in the Toussaint paper were not like this. In
> one case the neutrality (by having for example nine different codons
> for one amino acid) affected the effective mutation rate at different
> points in the genome. In the artificial case, an extra segment of the
> genome was not affecting the phenotype but did code for the mutation
> rate. In neither case does this do anything more than produce an
> accelerated learning algorithm, just like my
> neural-network-with-adaptive-learning-rates example. Neither example
> (as I recall from my brief reading of the paper) altered the size of
> the genome.

Correct. Size of the genome is another variable which if of real
interest. I am saying that neutrality is an essential component for
evolvability while duplication explains the scale free nature of
biological networks.

>
> As I said in the earlier post, gene duplication looks a more promising
> avenue if you're talking about design (or adaptation) of the fitness
> surface.
>
Gene duplication, given how common it seems to be, is indeed a promising
avenue as well but neutrality is a very important concept for evolvability.

[quote]Evolutionary algorithms apply the process of variation,
reproduction, and selection to look for an individual capable of solving
the task at hand. In order to improve the evolvability of a population
we propose to copy important characteristics of nature's search space.
Desired characteristics for a genotype-phenotype mapping are described
and several highly redundant genotype-phenotype mappings are analyzed in
the context of a population-based search. We show that evolvability,
defined as the ability of random variations to sometimes produce
improvement, is influenced by the existence of neutral networks in
genotype space. Redundant mappings allow the population to spread along
the network of neutral mutations and the population is quickly able to
recover after a change has occurred. The extent of the neutral networks
affects the interconnectivity of the search space and thereby affects
evolvability.[/quote]
How neutral networks influence evolvability Marc Ebner, Mark Shackleton,
Rob Shipman

See also
http://www2.informatik.uni-wuerzburg.de/staff/ebner/research/evolvability2/evolvability.html
for an example using pictures.

and

Wagner and Altenberg "complext adaptations and the evolution of
evolvability" 1996

    quote:
    ------------------------------------------------------------------------

    Abstract
    The problem of complex adaptations is studied in two largely
    disconnected research traditions: evolutionary biology and
    evolutionary computer science. This paper summarizes the results
    from both areas and compares their implications. In evolutionary
    computer science it was found that the Darwinian process of
    mutation, recombination and selection is not universally effective
    in improving complex systems like computer programs or chip designs.
    For adaptation to occur, these systems must possess "evolvability",
    i.e. the ability of random variations to sometimes produce
    improvement. It was found that evolvability critically depends on
    the way genetic variation maps onto phenotypic variation, an issue
    known as the representation problem. The genotype-phenotype map
    determines the variability of characters, which is the propensity to
    vary. Variability needs to be distinguished from variation, which
    are the actually realized differences between individuals. The
    genotype-phenotype map is the common theme underlying such varied
    biological phenomena as genetic canalization, developmental
    constraints, biological versatility, developmental dissociability,
    morphological integration, and many more. For evolutionary biology
    the representation problem has important implications: how is it
    that extant species acquired a genotype-phenotype map which allows
    improvement by mutation and selection? Is the genotype-phenotype map
    able to change in evolution? What are the selective forces, if any,
    that shape the genotype-phenotype map? We propose that the
    genotype-phenotype map can evolve by two main routes: epistatic
    mutations, or the creation of new genes. A common result for
    organismic design is modularity. By modularity we mean a
    genotype-phenotype map in which there are few pleiotropic effects
    among characters serving different functions, with pleiotropic
    effects falling mainly among characters that are part of a single
    functional complex. Such a design is expected to improve
    evolvability by limiting the interference between the adaptation of
    different functions. Several population genetic models are reviewed
    that are intended to explain the evolutionary origin of a modular
    design. While our current knowledge is insufficient to assess the
    plausibility of these models, they form the beginning of a framework
    for understanding the evolution of the genotype-phenotype map.

> I guess anyone on this list who is a Christian ultimately believes God
> designed it all - intended the universe to be so that intelligent
> creatures appeared who would be aware of their creator. Most people
> on this list would call themselves "Theistic Evolutionists", implying
> that God chose to create via the process of evolution. I would count
> myself in that camp these days.
>
> I'm wondering where you stand on this, Pim? You haven't said.

Same here. Which is why I find Intelligent Design so troublesome.
Received on Sat May 28 17:18:19 2005