Title : Structural phylogeny or the construction of gene history from protein structures.
Summary: As C. Chothia and A. Lesk demonstrated as early as 1986, protein structures, and more particularly protein folds, are better conserved during evolution than protein sequences. This property and the impressive accumulation of resolved protein structures now make it possible to enrich the concept of the 'structure-function relationship' and to add a 'structure-phylogeny relationship' component, with the aim of establishing more precise structural phylogenies than phylogenies based on sequence alignments. Two examples will be presented dealing with widespread canonical folding. The first is a family built on the repetition, duplication/fusion and reorganisation of a βαβββ module of about sixty amino acids. The second case presented will concern a module that is very widespread in the context of viral capsids: the structural domain known as jelly-roll. These two examples concern very different superfamilies but with a number of points in common: i) each is highly populated, ii) within each sequence similarities are almost undetectable, iii) folding is conserved and iv) each derives from a single ancestor following complex paths, histories.