Abstract

The informational properties of biological systems are the subject of much debate and research. I present a general argument in favor of the existence and central importance of information in organisms, followed by a case study of the genetic code and the translation system from the perspective of information. The codon biases of 831 Bacteria and Archeae are analyzed and modeled as points in a 64-dimensional statistical space. The major results are that codon bias evolution does not follow canonical patterns, and the use of coding space in organsims is a subset of the total possible coding space. These findings imply that codon bias is a unique adaptive mechanism that owes its existence to organisms' use of information in representing genes, and that there is a particularly biological character to the resulting biased coding and information use.