AVERAGE CORE STRUCTURES AND VARIABILITY MEASURES FOR PROTEIN FAMILIES- APPLICATION TO THE IMMUNOGLOBULINS

A variety of methods are currently available for creating multiple ali gnments, and these can be used to define and characterize families of related proteins, such as the globins or the immunoglobulins. We have developed a method for using a multiple alignment to identify an avera ge structural ''core'', a subset of atoms with low structural variatio n. We show how the means and variances of core-atom positions summariz e the commonalities and differences with a family, making them particu larly useful in compiling libraries of protein folds. We show further how it is possible to describe the rotation and translation relating t wo core structures, as in two domains of a multi-domain protein, in a consistent fashion in terms of a ''mean'' transformation and a deviati on about this mean. Once determined, our average core structures (with their implicit measure of structural variation) allow us to define a measure of structural similarity more informative than the usual root- mean-square (RMS) deviation in atomic position, i.e. a ''better RMS.'' Our average structures also permit straightforward comparisons betwee n variation in structure and sequence at each position in a family. We have applied our core-finding methodology in detail to the immunoglob ulin family We find that the structural variability we observe just wi thin the VL and VH domains anticipates the variability that others hav e observed throughout the whole immunoglobulin superfamily; that a cor e definition based on sequence conservation, somewhat surprisingly, do es not agree with one based on structural similarity; and that the cor es of the VL and VH domains vary about 5 degrees in relative orientati on across the known structures. (C) 1995 Academic Press Limited