On evolution / John Maynard Smith.

  • Maynard Smith, John, 1920-2004.
Date:
1972
Catalogue details

Licence: Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)

Credit: On evolution / John Maynard Smith. Source: Wellcome Collection.

    107/144 (page 95)
    Hence / must be greater than 1 /1000. It does not follow that the fraction of all possible sequences which are meaningful need to be as high as 1 /1000. It is probably much lower. Almost certainly, there is a higher probability that a sequence will be meaningful if it is a neighbour of an existing functional protein than if it is selected at random. No quantitative difficulty arises in explaining the evolution of proteins if/iV>l. The argument of course says nothing about the origin of life, since it assumes that at least one functional protein exists as a starting point. Before leaving this topic, it may be worth saying something about the geometry of the protein space ; these ideas emerged during a conversation with Donald Glaser. We want space in which two proteins are neighbours if they can be converted into one another by a single mutation. For simplicity, I will assume that amino acid substitutions are the only possible mutations and, ignoring the code, that all substitutions are possible. I will start with the space representing all possible dipeptides. This will be represented by a 20 x 20 chess board, each of the 400 squares representing a different peptide. A single mutational step is equivalent to a rook's move. Any dipeptide can be converted into any other by two moves. However such a conversion may be impermissible in evolution if the intermediate is meaningless. Suppose two squares ( dipeptides ) are connected by a meaningful path, how long can the shortest meaningful path between them be ? The answer is obviously the 38 rook moves required to travel from one corner to the opposite corner by moving one square alternately in a horizontal and vertical direction. This path would be shortened by the presence of any meaningful peptides other than those on the path. Transferring these ideas to the space of all proteins 100 amino acids long, the space would be of 100 dimensions, and contain 20i®o 'squares'. A mutation is still a rook's move, in any one of 100 directions. Any protein could be converted uito any other by not more than 100 mutations. If two proteins are connected by a meaningful path, the shortest meaningful path between them cannot be greater than 1900 steps, and will usually be much shorter. ( It was realized later that this is wrong. The maximum 'shortest meaningful path' could be longer than this by many orders of magnitude. Hence there may be large regions of the protein space as yet unexplored. ) 95