|
A given protein family can be divided into fundamental units of analysis called cohesion groups. Within a cohesion group, members are sufficiently close in evolutionary relationship to yield correct or nearly correct branching relationships on a phylogenetic tree (except when lateral gene transfer has occurred). As such, these are rigorous units for making bioinformatic and evolutionary inferences about various character states (e.g. substrate specificity) of the protein. Once the latter foundation is firmly in place, the scope of the analysis can be progressively enlarged with accuracy. The continual availability of new sequences is expected to result, not only in the temporary formulation of new cohesion groups, but also with the merging of cohesion groups as phylogenetic gaps are progressively filled. Concatenation of the seven enzymes of tryptophan biosynthesis has been shown to confer greatly expanded resolving power to a cohesion group. The assembly of such “supercohesion groups” is envisioned as an advanced step that will support ever more rigorous bioinformatic and evolutionary analysis. The cohesion-group approach, herein illustrated with the TyrA subsystem, promises enablement of a focused and correct knowledge base which is amenable to systematic and semi-automated expansion __ both within a metabolic subsystem and ultimately to connected subsystems of biochemical networks.
|
 |