A GENERALIZED INFORMATION FUNCTION APPLIED TO THE GENETIC-CODE

The problem of the partitioning of the degeneracy of the codons in the genetic code is considered in the framework of a generalized information function I(G) = c-SIGMA(k)-p(k){ln p(k) + G(E(k))} where k represents the number of condons in a specific degeneracy class and G(E(k)) is an arbitrary real valued function. For G(E(k)) = 0 the Shannon information function is recovered. For a particular choice of G(E(k)) that takes the dominance of even degeneracies into account, it is found by direct numerical calculations that the correct degeneracy partitioning appears as optimal values of the I(g) function. This result is also supported by optimization calculations in which the generalized information function is regarded as a continuous function in the degeneracy variables.