The evolution of signaling complexity suggests a mechanism for reducing the genomic search space in human association...

The size complexity of the human genome has been traditionally viewed as an obstacle that frustrates efforts aimed at identifying the genetic correlates of complex human phenotypes. As such complex phenotypes are attributed to the combined action of numerous genomic loci, attempts to identify the underlying multi-locus interactions may produce a combinatorial sum of false positives that drown out the real signal. Faced with such grim prospects for successfully identifying the genetic basis of complex phenotypes, many geneticists simply disregard epistatic interactions altogether. However, the emerging picture from systems biology is that the cellular programs encoded by the genome utilize nested signaling hierarchies to integrate a number of loosely coupled, semiautonomous, and functionally distinct genetic networks. The current view of these modules is that connections encoding inter-module signaling are relatively sparse, while the gene-to-gene (protein-to-protein) interactions within a particular module are typically denser. We believe that each of these modules is encoded by a finite set of discontinuous, sequence-specific, genomic intervals that are functionally linked to association rules, which correlate directly to features in the environment. Furthermore, because these environmental association rules have evolved incrementally over time, we explore theoretical models of cellular evolution to better understand the role of evolution in genomic complexity. Specifically, we present a conceptual framework for (1) reducing genomic complexity by partitioning the genome into subsets composed of functionally distinct genetic modules and (2) improving the selection of coding region SNPs, which results in an increased probability of identifying functionally relevant SNPs. Additionally, we introduce the notion of 'genomic closure,' which provides a quantitative measure of how functionally insulated a specific genetic module might be from the influence of the rest of the genome. We suggest that the development and use of theoretical models can provide insight into the nature of biological systems and may lead to significant improvements in computational algorithms designed to reduce the complexity of the human genome.