Minimizing catastrophic mutations as the selective force in the evolution of the universal modern genetic code.
An early addition of a Stop signal to the first functional genetic codes robustly predicts the subsequent evolution of codon to amino acid pairings in the universal modern genetic code.
Recent Comments
Peter, could the next test distinguish a code’s resistance to harmful mutations from the probability of evolving it?
In your expansion scenario, previously uninformative codon positions acquire specificity while established assignments are retained. The question is how those steps occur through changes in decoding, including any temporary ambiguity.
Could a finite population model test whether early Stop signals make the predicted Stop/cysteine arrangement more likely across these expansion paths, allowing genetic drift and uncertainty in amino acid availability?
I would compare how often that arrangement emerges and where expansion stalls, using the same transition rules with and without the proposed catastrophic mutation costs.
That would test whether the mechanism helps generate the pattern, beyond making the reconstructed pattern comparatively low in mutation cost. Which expansion step would you expect to be most restrictive?
Models of the evolution of the code are kind of stuck with two facts that simulations are trying to match and explain: The observed pairings in the modern code, and the most likely order in which AA were added. If there is no negative pole (a catastrophic stop codon) to avoid, then putting them here is probably no better or worse than putting them there. In short, there are probably a lot of alternatives that would be as good or maybe just a bit better than the modern arrangement. Also, there is the question of how are the codons functioning. The argument seem to be strong that they were always in triplets. Does that mean the earliest codes were massively synonymous at the 1st and 3rd letter, or were they just meaningless spacers? That would kind of affect where is the best place to add (if there is no stop to worry about). To answer your question, it seems analogous to asking which square is the worst to make your first move in a game of tic-tac-toe. The answer is there is no worst for you X, the outcome depends on what the next moves are. There very much is a 'worst' for the second O move!
I think of the various assumptions we make, the easiest to defend is that a mutation to stop is generally worse than a change of AA. Thus, when you ask about a no catastrophe scenario, that would most likely mean no stops in the code. That's hard to simulate because what do you put at the UAx codons? There is no obvious other AA candidate to stick there. If stop is the last thing added (as some argue), it seems a bit far fetched that UA remains empty until the very end - waiting for that stop!
Overall, the pattern we see is that the code evolved to seemingly steer clear of the UA codon (and to smaller degree the UG after cysteine) as AA are added. If not stops, what else logically could produce this effect?