Semantic closure

Our paper “Semantic closure demonstrated by the evolution of a universal constructor architecture in an artificial chemistry” has just been published in the Journal of the Royal Society Interface.  We submitted in December, it was accepted after revision on 24 April, and appeared online yesterday! The advantages of web-based publishing.

Our “media friendly” summary is:

The ‘meaning’ of DNA lies in the act of translating a DNA sequence into a protein sequence. The mapping of DNA to proteins is identical in nearly all species, but some species have evolved alternative mappings. A new computer model uses an artificial chemistry to investigate evolutionary changes in these mappings, where the translating apparatus is encoded in the DNA and governs its own translation. As well as reproducing the known evolutionary mechanism of changing the meaning of DNA, the model predicts a novel mechanism for changing the mapping in biology that is not detectable by phylogenetic DNA sequence analysis.

Our slightly less friendly paper abstract is:

Abstract: We present a novel stringmol-based artificial chemistry system modelled on the universal constructor architecture (UCA) first explored by von Neumann. In a UCA, machines interact with an abstract description of themselves to replicate by copying the abstract description and constructing the machines that the abstract description encodes. DNA-based replication follows this architecture, with DNA being the abstract description, the polymerase being the copier, and the ribosome being the principal machine in expressing what is encoded on the DNA. This architecture is semantically closed as the machine that defines what the abstract description means is itself encoded on that abstract description.We present a series of experiments with the stringmol UCA that show the evolution of the meaning of genomic material, allowing the concept of semantic closure and transitions between semantically closed states to be elucidated in the light of concrete examples. We present results where, for the first time in an in silico system, simultaneous evolution of the genomic material, copier and constructor of a UCA, giving rise to viable offspring.

This is one of the key findings:

Figure 6. Semantic change without mutation of the genome.

Genome G0 (built in our artificial chemistry StringMol) encodes a bunch of “machines”, including E0.  E0 reads G0 and expresses the machines encoded on it.  The expression processes can make mistakes: one such mistake meant that E0 expressed E1 instead of another E0.  This “mutant” machine E1 then expressed E2 (without error).  And then E2 expressed itself, again without error. So the meaning of that part of the genome where the expressor is encoded has changed from E0 to E2.  All without the genome changing.  Which is cool.

The paper is open access and can be found at doi:10.1098/rsif.2016.1033.