The Optimal Design of the Genetic Code

@NLENTS Faz Rana seems to beleive that the genetic code was an optimal design. Your book “Human Errors” takes the exact opposite view. Any comments on the article below. As always everyone’s comments are welcomed.

1 Like

Is there some reason that evolution cannot find the optimal design for a system? Put another way, why wouldn’t we expect to see optimal designs from evolutionary processes? Is it just too improbable?

How could we know what optimal is in this scenario? Optimal with respect to what? How do we know this is the normative criteria versus another?

Also the question depends on what appears to be a factual error. There is no universal genetic code. We see several genetic codes in nature, and even in the same organism. Humans, for example, have two genetic codes, one for their nuclear DNA and one for their mitochondria.

1 Like

The genetic code does seem pretty darn optimal. @Zachary_Ardern has covered some of this. He links the relevant work in this piece:

I enjoy this one:

1 Like

By what criteria?

This subject was discussed on this very forum last year:

@vjtorley referenced this interesting paper by Eugene Koonin:

1 Like

Are you familar with the work of Michael Yarus and Bojan Zagrovic? @evograd

The names ring a bell, but I couldn’t describe their work off the top of my head.

Koonin mentions Yarus rather briefly in that paper. But what they have shown is there are chemical affinities between amino acids and their codons and anti codons. Here’s a little bit on Zagrovic:

My blogpost listed five properties of the standard genetic code (the other codes are all derivative of it and fairly minor in the greater scheme of things as far as I’m aware) in which respectable published papers have claimed the code to be optimized i.e. much better than most possible alternative codes. There are other claims for optimized properties out there too, which I haven’t sifted through carefully enough.
The proper comparison set though is debatable, and there are various accounts of the evolutionary origin of the code - I have no problem with such accounts in principle or theologically or anything like that, but they are highly speculative, and none that I’m aware of consider the multiple optimalities observed. More work on modelling code origin stories is needed I think, but it seems to me that the available data suggests we were very lucky indeed to get the code we did.
As the possibility space is ginormous (as in some other important areas of ev. bio) it may be impossible to satisfactorily model code evolution in order to work out just how lucky we are with much confidence.


Right, yeah, I’ve been intrigued by that.

I should also add that these are two contradictory claims:

  1. The genetic code is optimal, with the ideal assignment of codons to amino acids.
  2. The genetic code is a true code, with arbitrary assignment of codons to amino acids.

ID people make both arguments as evidence of design. Which one is it? You can’t argue both…

Hi Zach - I haven’t seen your blog post yet, but the “Texas Sharpshooter Problem” would seem to be a huge difficulty. I’ll look for the post when I have a chance.

Are they? They seem to be arguing we have the optimal code but there is no reason why certain codons code for certain amino acids. So why this code if there was no physical necessity? What am I missing?


Current Evolutionary theory is a “come ss you are party”. You can only expect so much “beauty” when ones genetic options are confined to an ancestral population… and some mutations.

However, if we bring God into the picture (Evolution, God-Governed, or E.G.G.) … we are forced to confess that even with God’s design… some of his designs are way better than others employed for other lineages.

It brings us right to the limits of what we think we can understand about God and His/Her plan!

Wouldnt you, too, agree with this, @Ashwin_s?

From my semi-reading of the paper, the main point seems to be focused on the impact of mutations on protein function. The genetic code seems to favor conservative changes in the peptide sequence for many mutations. In other words, mutations in a codon for a hydrophobic amino acid will tend to result in another hydrophobic amino acid. This can help reduce the number of detrimental mutations.

Using conservation of amino acid properties as their criteria they were able to look at other possible combinations of codons and amino acids, and they were able to find better genetic codes. I think this points to optimization of an ancestral genetic code since it mimics the famous fitness landscape that @swamidass linked to in a previous post.

Seems that the genetic code is not the only optimal design that evolution is capable of bringing about!

Magnetic Compass of Birds Is Based on a Molecule with Optimal Directional Sensitivity