6.4  Monday, Jan. 4  Protein Modularity and Evolvability: Evolutionary Origins and Consequences RORICK, Mary/M.*; WAGNER, Gunter/P.; Yale University, New Haven, CT mary.rorick@yale.edu

It is well recognized that “modularity” is an important feature of living systems, though we lack concrete definitions for what this term means. Being able to quantify modularity, at least at one level of biological hierarchy, is essential for addressing questions about its evolutionary origins and implications. Here we develop an index for protein modularity that reflects how much of the protein’s tertiary structure is constrained by coevolving amino acids. A quantitative method for measuring protein modularity makes it possible to test whether there is natural variation in modularity, and whether it correlates with directional selection. If we find a correlation between modularity and directional selection at the protein level, we can conclude that biological modularity is the product of natural selection for evolvability.

We test for this correlation in a dataset of orthologous genomic markers that have solved tertiary structures. For each protein in the dataset, we use a site model to obtain a proportion of sites under directional selection, p, and an associated ω for this category of sites. Our index for directional selection is p(ω-1). Our index for modularity is the P-value of the average distance between coevolving pairs of amino acids in the tertiary structure.

Given that proteins are the fundamental building blocks of essentially all biological systems, addressing whether protein modularity evolves for the sake of protein evolvability is its own justification. However, to the extent that our indexes are grounded in formal definitions that are independent of the specific context within biological hierarchy, it may be possible to extrapolate from our findings to form hypotheses about modularity and evolutionary dynamics at other biological levels, such that of the whole organism.