On the Origin of Macromolecular Sequences

Abstract

The origin of the degree and type of order found in biological macromolecules is not adequately explained solely as an accumulation of genetic restrictions acquired through natural selection from otherwise unrestricted primeval sequences capable of self-replication, since the biological process of replication is itself dependent on the pre-existence of such order, and since the number of sequences that could have ever been tested by selection on the earth is an insignificant fraction of the number of unrestricted sequences which would be possible. Therefore the hypothesis is considered that replication and selection began from well ordered sequences, rather than random sequences. It is shown how the Turing concept of computation in fed-back, discrete-state automata can lead to the generation of order withour pre-existing instructions, and how this computation can result in self-repeating, random-like, but well ordered sequences of great length. Macromolecular models of such computers are suggested on the basis of mechanisms proposed for the growth of eutactic polymers. Such self-replicating, mutable sequences may then evolve genetic control which is sufficient to accommodate the information accumulated by natural selection. The structure and function of enzymes and structural proteins is related to this model, and statistical evidence from known amino acid sequences is shown to be consistent with some degree of non-genetic ordering.