Evolution as a communication channel

We present a mathematical framework based on the Information Theory of Claude Shannon for modelling the evolution as an information transmission process. The evolution is visualised as a communication system in which the message being sent is the genetic code. Due to the mutations, errors are introduced on the message sent. These errors define the noise in the communication system analogy. Using various substitution matrix models in the literature we calculate the capacity of the evolution channel for both non-coding RNA and the protein coding DNA. We calculate the capacities also based on experimental data for both types of channels. The capacities calculated give us the theoretical limits of information transferable via genetic code without error and hence the number of generations (or unit time) required for the alteration of the sent message. The difference between protein coding and non-coding channels gives us the measure of the robustness to mutations provided by the protein code.