The information capacity of the genetic code: Is the natural code optimal?

• Published in: Journal of theoretical biology Vol. 419; pp. 227 - 237
• Main Authors: Kuruoglu, Ercan E., Arndt, Peter F.
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 21.04.2017
• Subjects: Algorithms; Amino Acids - genetics; Base Sequence; Codon - genetics; DNA; Evolution, Molecular; Genetic code; Genetic Code - genetics; Information capacity; Information theory; Models, Genetic; Shannon theory; Genetic code; Information capacity; Shannon theory; DNA; Information theory
• ISSN: 0022-5193; 1095-8541; 1095-8541
• DOI: 10.1016/j.jtbi.2017.01.046

We envision the molecular evolution process as an information transfer process and provide a quantitative measure for information preservation in terms of the channel capacity according to the channel coding theorem of Shannon. We calculate Information capacities of DNA on the nucleotide (for non-coding DNA) and the amino acid (for coding DNA) level using various substitution models. We extend our results on coding DNA to a discussion about the optimality of the natural codon-amino acid code. We provide the results of an adaptive search algorithm in the code domain and demonstrate the existence of a large number of genetic codes with higher information capacity. Our results support the hypothesis of an ancient extension from a 2-nucleotide codon to the current 3-nucleotide codon code to encode the various amino acids. •Study of evolution of protein coding modelled as a communication system.•study of the optimality of the natural genetic code with channel capacity as objective function.•an intelligent algorithm to search optimal genetic codes, making the largest search up to date.•demonstrating that the natural genetic code is suboptimal despite being one in a million.•observations supporting the hypothesis that previously codons were composed of 2 nucleotides.