the genetic code

DNA and RNA are written in an alphabet of just four letters — the bases A, U (or T), G and C — yet proteins are built from twenty different amino acids. So there is a translation problem: how do you spell out twenty things with only four letters? The genetic code is the dictionary that solves it. It is the set of rules that says which run of RNA letters stands for which amino acid.

The code works in three-letter words. Each group of three consecutive bases in a messenger RNA is a codon, and each codon names one amino acid (or a 'stop' signal). Three letters drawn from four possibilities gives 4 x 4 x 4 = 64 possible codons — comfortably more than the twenty amino acids needed, so most amino acids are spelled by more than one codon. One codon, AUG, does double duty as the universal 'start' that also codes for methionine, and three codons (UAA, UAG, UGA) spell no amino acid at all but mean 'stop here'. The code is read strictly in order, never skipping or overlapping, from a fixed starting point — so the same string of letters means something completely different if you begin reading one base too early or too late.

This code is the hinge of the central dogma's last arrow, RNA -> protein. What makes it remarkable is its near-universality: a bacterium, a banana, a blue whale and you all use essentially the same dictionary, which is why a human gene can be put into bacteria and still make the right protein — the foundation of the entire biotechnology industry. The code is not a law of chemistry, though; there is no physical reason CUG 'must' mean leucine. It is a frozen historical accident, locked in so deep that life has barely been able to change it in billions of years.