We previously showed that genomic structure is constructed homogeneously with putative small units coding similar amino acid composition. The present study investigates the relationship between protein and codon formation from genetic structures, assuming that a structure can sometimes reveal its formation process. Genes are apparently constructed with roughly similar small repeating units encoding about 100 amino acid residues in respect of the amino acid composition and codon usage, even though the amino acid sequences of the small units are completely different. During prebiotic evolution, polymerizations of amino acids or nucleotides occurred initially in the absence of a codon system. To simulate this initial polymerization, we applied a simulation analysis based on a random choice of monomer, amino acid or nucleotide. In genes, the amino acid compositions presumed from the units encoding 58-464 amino acid residues exactly fit those obtained from a simulation analysis based on a random choice of amino acid from an amino acid pool. However, the amino acid compositions of the small units do not fit the amino acid compositions similarly calculated, based on a random choice of nucleotide from the nucleotide pool. Thus, to result in the present genetic structure, codon formation must have been controlled under certain conditions. This suggests that, mathematically, protein formation chronologically preceded that of the codon formation during the evolution of primitive life.
Buy this Article