Emergence of algorithmic language in genetic systems

Area of Science:

• Computational Biology
• Evolutionary Computation
• Systems Biology

Background:

• The genotype-phenotype interface is complex, akin to a computer where chromosomes are algorithms and phenotypes are computations.
• Finding functional algorithms via random mutation (brittleness problem) is challenging in genetic systems.
• Previous models often simplify the genotype-phenotype mapping, limiting biological realism.

Purpose of the Study:

• To investigate how evolutionary operators like mutation and crossover influence the emergence of an "algorithmic language" in genetic systems.
• To demonstrate that this emergent language can mitigate the brittleness problem, facilitating the generation of meaningful phenotypes.
• To analyze the population dynamics of a neurogenetic model to understand language emergence and its role in solving evolutionary challenges.

Main Methods:

• Utilized a variant of Kitano's neurogenetic model, simulating population dynamics.
• Represented chromosomes as encoding rules for cellular division.
• Interpreted the resulting 16-cell organism phenotype as a connectivity matrix for a feed-forward neural network.

Main Results:

• Observed the emergence of a structured "algorithmic language" within the genotype.
• Detailed the characteristics and grammar of this emergent language.
• Demonstrated a significant reduction in the "brittleness problem" due to the facilitated production of meaningful sequences.

Conclusions:

• Evolutionary processes, specifically mutation and crossover, can drive the emergence of sophisticated algorithmic languages.
• This emergent language acts as a crucial intermediary, simplifying the genotype-phenotype mapping and overcoming evolutionary hurdles.
• The findings provide insights into the evolution of complexity and the robustness of genetic systems.