Chapter category: Gene Expression
Birth and Death Models of Genome Evolution
Chapter authors:
Georgy P. Karev, Yuri I. Wolf, and Eugene V. Koonin
Gene duplication is the primary avenue of genome evolution. The gene repertoire of any species can be described as an ensemble of paralogous gene families, ranging in size from one to large numbers that amount to a substantial fraction of genes in the respective genome. Evolution of such an ensemble is naturally represented by a birth-and-death process, the birth of a gene being duplication, and death being gene inactivation and elimination. In addition to gene duplication and loss, evolution of gene families involves “true” innovation, i. e., appearance of genes new to the given lineage through horizontal gene transfer, emergence of genes from noncoding sequences, and change of preexisting genes beyond recognition. Assuming these three elementary processes, we developed a simple theoretical framework for analysis of genome evolution, the Birth, Death and Innovation Models (BDIMs). Comparison of the predictions made by different versions of BDIMs with empirical distributions of paralogous family size in genomes allows one to choose the adequate models. Stable family size distributions can evolve only under balanced BDIMs, in which duplication and deletion rates are asymptotically equal up to the second order. The linear BDIM, in which there is almost no dependence between the family size and birth-death rates, readily approximates the observed family size distribution at equilibrium. However, the stochastic version of this model yields unrealistic times for evolution of the large paralogous families that were detected in all genomes. In order to produce reasonable rates of family evolution, one needs to turn to nonlinear higher-degree BDIMs, which imply “interactions” between paralogs. These interactions may be interpreted as a proxy for natural selection, which should drive evolution of large paralogous families if their emergence is to be viewed as an adaptive reaction.
