Synonymous substitution

Since there are 22 codes for 64 codons, roughly we should expect a random substitution to be synonymous with probability about 22/64 = 34%. The actual value is around 20%.

A synonymous mutation can affect transcription, splicing, mRNA transport, and translation, any of which could alter the resulting phenotype, rendering the synonymous mutation non-silent. The substrate specificity of the tRNA to the rare codon can affect the timing of translation, and in turn the co-translational folding of the protein. This is reflected in the codon usage bias that is observed in many species. A nonsynonymous substitution results in a change in amino acid that may be arbitrarily further classified as conservative (a change to an amino acid with similar physiochemical properties), semi-conservative (e.g. negatively to positively charged amino acid), or radical (vastly different amino acid).

Degeneracy of the genetic code

Protein translation involves a set of twenty amino acids. Each of these amino acids is coded for by a sequence of three DNA base pairs called a codon. Because there are 64 possible codons, but only 20-22 encoded amino acids (in nature) and a stop signal (i.e. up to three codons that do not code for any amino acid and are known as stop codons, indicating that translation should stop), some amino acids are coded for by 2, 3, 4, or 6 different codons. For example, the codons TTT and TTC both code for the amino acid phenylalanine. This is often referred to as redundancy of the genetic code. There are two mechanisms for redundancy: several different transfer RNAs can deliver the same amino acid, or one tRNA can have a non-standard wobble base in position three of the anti-codon, which recognises more than one base in the codon.

In the above phenylalanine example, suppose that the base in position 3 of a TTT codon got substituted to a C, leaving the codon TTC. The amino acid at that position in the protein will remain a phenylalanine. Hence, the substitution is a synonymous one.