transition versus transversion

When one DNA base is swapped for another in a point mutation, the swap comes in two flavours, and which flavour it is depends on the chemistry of the bases involved. The four bases fall into two families by shape: the purines (A and G) are the larger, double-ring bases, and the pyrimidines (C and T) are the smaller, single-ring bases. A transition keeps within a family — a purine becomes the other purine, or a pyrimidine becomes the other pyrimidine. A transversion crosses between families — a purine turns into a pyrimidine or vice versa.

So A-to-G and C-to-T are transitions (like-for-like in ring shape), while A-to-T or G-to-C are transversions (a big ring traded for a small one or the reverse). There are only four possible transitions but eight possible transversions, yet — surprisingly — transitions occur far more often in real genomes. The chemistry favours them: a very common spontaneous event, the deamination of cytosine to uracil (read as T), drives C-to-T transitions, and the copying machinery tends to slip more easily between same-shaped bases.

The distinction is more than bookkeeping. The ratio of transitions to transversions is a standard quality check on DNA sequencing data and a clue in evolutionary studies, because real biological variation shows a characteristic bias toward transitions. Transitions are also gentler on proteins on average: because of how the genetic code is laid out, a transition in the third position of a codon is more likely to leave the encoded amino acid unchanged.