http://people.csail.mit.edu/jaffer/GTOL/breakthrough
	Big Game Theory -- Evolutionary Breakthroughs

But ribosomal-RNA, a genetic molecule capable of synthesizing an infinite variety of non-genetic polymer molecules, seems fantastically complex.

Evolution in the Nucleotide Game argues that the first synthesized non-genetic polymers to evolve will be capsid components. Needing only to be uniform sized, self-assembling, easily disassembled, and more rugged than genetic polymers, they should require far fewer species of non-genetic monomers than later evolved enzymes.

The easy disassembly and uniform size requirements rule out a machine which just churns out polymers of a single monomer (or simple repeating sequence). It is hard to imagine how a simple counter which cut off synthesis after the correct number of repetitions would evolve from simple mutations. And a counter would be brittle, the vast majority of mutations causing it to be nonfunctional.

A more likely solution is found in the Turing machine, which counts using its tape. Genetic sequencing provides not only a count of polymer length, but arbitrary sequences of the monomers it encodes. For the purposes of synthesizing the non-genetic molecules, the effective length and content of the genetic sequence is frequently altered by mutations in the genetic material; providing variations of phenotype which spur evolution.

This first translation machine, while an order of magnituded simpler than ribosomal-RNA, would still require a great many mutations to an initial viroid. These mutations need not be simultaneous. Evolution in the Nucleotide Game argues that binding of genetic sites to non-genetic molecules would evolve to take advantage of their chemical properties. Fostering reactions between some of those molecules would also be advantageous; as would release of some molecules.

Some fragments of genetic material which bind to non-genetic molecules would have exposed nucleotides, allowing them to bind to genetic material. Thus transfer-RNAs would arise naturally, whether or not they are required for synthesis.

When viewed as steps, the development of translation machines is more plausible than it might have seemed at the outset. Different translation molecules would evolve simultaneously. Being able to synthesize from more than one monomer imparts a large adaptive advantage; as simple genetic mutations cause the creation of new varieties of synthesized molecules.

Genetic material which doesn't code for capsids will also be translated. Some of these molecules will be active chemically; and a few will precipitate useful reactions. Those genes will be seleted for and evolve along with the translation machine. Thus enzymes are born.

More than one type of translation machine will be incorporated into some evolving viruses. But cross synthesis from genetic material will likely waste resources, reducing its fitness. Recombination of genetic material may evolve combined translation machines. But even in the absence of recombination, the advantages accruing from incremental inclusion of new monomers types will ensure that translation machines evolve to be general.

These steps are all plausible results of mutations; but there are many steps. This process could be expected to take an eon. The game is long, but the rewards are enormous. As the translation machines improve to synthesize from more and more molecules they cross a threshold marking a qualitative distinction between this big game and the early nucleotide games. Having both protective capsids and a wide variety of enzymes imparts such a large competitive advantage that these players win the competition with the lagging ones. Because of interspecies exchange of genetic material, more than one species can emerge with this evolutionary breakthrough advantage.

Although the first steps toward translation could evolve in viroids after this, they would need to be isolated in a nutritive environment for an eon to complete it. Thus they would lose to advanced, synthesizing players if there was any contact during that time. Other big games in Earth history were the development of photosynthesis and aerobic respiration. Those breakthroughs will be adapted, but not recreated. An example of a "small game" which can evolve multiple times is explored in Viruses and Bacteria -- Flagellum.

Conclusions

• Complicated processes like protein synthesis, photosynthesis, and aerobic respiration culminate from mutations over very long periods of time.

• The evolutionary steps leading to complicated processes will often involve molecules used for unrelated processes.

• Because of the long time period to evolve a complicated process, it will re-evolve only in a large environment isolated for an eon from successful organisms incorporating the process.

• Minor improvements and variations to complicated processes can evolve even after the process establishes its dominance.