The earliest organisms are thought to have had high mutation rates. It has been asserted that these high mutation rates would have severely limited the information content of early genomes. This has led to a well-known "paradox" because, in contemporary organisms, the mechanisms that suppress mutations are quite complex and a substantial amount of information is required to construct these mechanisms. The paradox arises because it is not clear how efficient error-suppressing mechanisms could have evolved, and thus allowed the evolution of complex organisms, at a time when mutation rates were too high to permit the maintenance of very substantial amounts of information within genomes. Here, we use concepts from the formal theory of information to calculate the amount of genomic information that can be maintained. We identify conditions under which much higher levels of genomic information can be maintained than previously considered possible among origin-of-life researchers. In particular, we find that the highest levels of information are maintained when many genotypes produce identical phenotypes, and when reproduction occasionally involves recombination between multiple parental genomes. There is a good reason to believe that these conditions are relevant for very early organisms, and thus the results presented may provide a solution to a long-standing logical problem associated with the early evolution of life.
K E Y W O R D S :Adaptation, epistasis, models/simulations, mutations, population genetics, sex.Eigen's Paradox is a well-known logical problem associated with the origin of complex organisms (Eigen 1971;Eigen and Schuster 1979;Maynard Smith and Szathmáry 1995). Experimental data and logical considerations have led origin-of-life researchers to believe that, early in the history of life, mutation rates were much higher than they are in contemporary organisms (Eigen 1971;Eigen and Schuster 1979;Maynard Smith and Szathmáry 1995). According to Eigen and Schuster, this implies that the maximum amount of information that could have been stably encoded in the genomes of early organisms must have been severely limited (Eigen 1971;Eigen and Schuster 1979). In contemporary organisms, the mechanisms of error prevention and correction are quite complex. This leads to a "chicken-and-egg problem." How could life that is complex enough to suppress mutation to low levels have evolved while mutation rates were quite high?Eigen and Schuster's calculations are based on the idea that if the genome with the best-possible fitness cannot be maintained in a population, then "The information . . . would slowly seep away until it is entirely lost" (Eigen and Schuster 1979). Using this idea, Eigen and Schuster claimed that, for realistic parameter values, a meaningful genetic sequence cannot include much more than approximately 1/μ nucleotides, where μ is the pernucleotide mutation rate (Eigen and Schuster 1979). Assuming four equally likely nucleotides, it requires 2 bits of information to specify each nucleotide. Thus, Eigen ...