| One of the key unsolved problems in Biology today is understanding impact of classical evolution, on organismal and population level on molecular evolution of genes and proteins. We noted that evolution of populations of organisms each carrying M genes is isomorphic to the problem of M-dimensional random walkers in space of protein sequences with two adsorbing boundary conditions: at high energies of protein native conformations where proteins become unviable and organisms carrying their genes die and at lower energies where proteins are depleted of possible stable sequences. This problem can be solved exactly and we obtained the relation between mutation rates, duplication rate and stability range of the proteins at which populations remain viable. This formula predicts the effect of mutational meltdown and provides important insights into genomic organization and evolution of viruses and simple organisms. It predicts that RNA viruses are much shorter than dsDNA ones and that genomes of thermophilic bacteria are shorter than that of their mesophilic counterparts. All these predictions are verified in bioinformatics analyses and in mutational experiments on RNA viruses. Next, I will present a new microscopic multiscale model of evolution and adaptation in living cells, which couples population genetics with exact model of protein folding and protein-protein interactions. In this model properties of proteins and fitness of organisms are derived directly from their genomic sequences. We found that mutation rate is a selectable trait whereby adaptation requires transient raise of mutators with subsequent annealing to lower rates. The fitness landscape in this realistic model is very complex giving rise to rich biological behavior. In light of these results some popular population genetics concepts such as single fitness peak should be considerably revised. |
|