"The recent excitement in the coding problem"

Date:: 1963

Reference:: PP/CRI/I/1/22

Catalogue details

Licence: In copyright

Credit: "The recent excitement in the coding problem". Source: Wellcome Collection.

6/56

codon for another amino acid, then it is called mis-sense. If it is changed to a codon that does not stand for an amino acid, it is called nonsense. If such a change completely destroys the function of the gene, the codon is called absolute nonsense. It is not yet known whether all nonsense is absolute nonsense. 9. Ambiguous and Unique A codon that stands at the same time for more than one amino acid is said to be ambiguous. A code, all the codons of which are unambiguous, is said to be unique. If some of the codons are ambiguous, the code is said to be partly ambiguous. II. General Questions A. Is the Code Overlapping? The first fairly precise code, suggested by Gamow, was of the over lapping type. Gamow (1954) argued that this was plausible for stereo chemical reasons. Adjacent amino acids in a polypeptide chain can be no more than 3.7 Ä apart, so that, in order to be joined together, they must be brought fairly close to each other. Adjacent nucleotides are about the same sort of distance apart—the exact distance depends upon which part of the nucleotide is measured. The bases tend to stack with their planes 3.4Â apart, but two adjacent phosphates (or two sugars) are usually 5-6 A apart. However, for a nonoverlapping triplet code the relevant distance is not between adjacent nucleotides but between every third nucleotide. In the usual sort of structure this distance is likely to be 10 Â, or greater. A simple picture of the assembly process would place adjacent amino acids too far apart to be joined together. We realize now that there are several other ways of getting round this difficulty but at the time it seemed that a fully overlapping code would be rather a neat solution. Any fully overlapping code will impose restrictions on the amino acid sequence, since the linear density of information is not the same in the two languages, there being at each step only four choices for a base, but potentially twenty choices for an amino acid. Thus the latter choice must necessarily be restricted. It was easy to disprove the code actually suggested by Gamow. This was of the overlapping triplet type and it was fully and logically degenerate. Apart from the fact that it failed to specify the direction of the polypeptide chain, it could be shown that it would not code for the known sequence of insulin. A search was next made to see if there were any restrictions on amino