Metabolism

From WikiManual
Revision as of 20:30, 14 February 2014 by PhiNotPi (talk | contribs) (Fixed typos)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Metabolism in Darwinbots is still in the planning stages. The exact implementation would rely heavily on technical constraints, specifically the number of different chemicals the program can support with reasonable speed and memory issues.

Chemicals

Each chemical in the program is considered to be a simple organic molecule derived from a relatively small collection of groups (analogous to the hydroxyl etc. groups in biochemistry).

Ideally there are perhaps 8 groups, with each molecule being represented by the same number of bits. A bit of 0 would represent the absence of that group, whereas a bit of 1 would represent the presence of that group.

For example:

10001000 would indicate a molecule made up of group 1 and 5.

00000000 would indicate a molecule made up of none of the groups (but presumably still possessing the "back bone" of all the other organic molecules, and so still having chemical properties, mass, volume, etc.) This molecule would be an acceptor for all groups.

Each chemical is assigned arbitrary attributes determining free nrg, shell, slime, poison, etc.

Metabolic Paths and Enzymes

All metabolic paths follow this form:

A + Enzyme (+/- nrg) -> B + Changed Enzyme (+/- nrg)

Each enzyme has a single (or possibly more) group it facilitates the transfer of. An enzyme that has a group attached to it is "full" and an enzyme without a group is "empty". It is assumed that any chemical with the group in question is a valid donor and thus can donate its group to an "empty" enzyme. Any chemical without the group is a valid acceptor, and can receive the group from a "full" enzyme. Enzymes can become more specific over time (using methods I haven't really examined yet) to accept and donate to specific chemicals.

Pumps (ie: Transporters)

To facilitate the movement of substances from and to distinct areas (say, from outside the cell to inside the cell), 'pumps' are formed. Pumps are similar to enzymes. Pumps can specialize/despecialize through mutations.

Each pump moves substances from one area to another. This pumping action requires nrg (or perhaps can generate nrg, more on this in a second). Pumps are ambi-directional depending on the needs of the cell (can act as an exit or an entrance), and (perhaps) may be used to generate nrg similar to the proton pumps in actual biological cells (this would actually be quite complex and is really its own separate idea).

Generating nrg would probably involve taking advantage of osmotic forces.

Mutations

Taken Directly from the PLoS article listed in sources:

We assume that initially all enzymes and all transporters are unspecific and allow small mutations in their kinetic properties. These mutations increase the rate constant of a single reaction while decreasing the rate constants of all other reactions of an enzyme or transporter. Alternatively they decrease the rate constant of a single reaction while increasing the rate constants of the other reactions. In addition to mutations affecting kinetic properties, we assume that there are gene duplications (or deletions) that increase (decrease) the dosage of enzymes and transporters.

Changing the dosage of an enzyme has two opposing effects on fitness: An increased dosage increases the fluxes in the network, but it is also associated with increased metabolic costs of biomass formation. Duplications also enable enzymes to diverge and specialize on different reactions. Note that if a multifunctional enzyme has only a small impact on biomass formation, duplication can be detrimental provided that the metabolic costs of the additional copy are larger than the benefits in terms of increased growth rate. Consequently, enzymes and transporters do not necessarily specialize completely over the course of evolution. On the other hand, duplications can be beneficial if enzymes or transporters are already completely specialized. Therefore multiple identical copies of completely specialized enzymes and transporters can evolve provided they have a large impact on biomass formation.

Final Thoughts

The ideas expressed here are not set in stone by any means. If you feel you know a better way, feel free to discuss it here or in the forums.

Sources

The Evolution of Connectivity in Metabolic Networks