Frozen Evolution. Or, that’s not the way it is, Mr. Darwin. A Farewell to Selfish Gene.

- The term selfish DNA, i.e. designation for DNA segments that proliferate in the gene pool through molecular drive, must not be confused with the similar, but totally unrelated, term selfish gene or the somewhat related term ultraselfish gene.The selfish gene is the central concept of Dawkin’s model of biological evolution (Dawkins 1976)This model, based on the theoretical work of W.D. Hamilton (Hamilton 1964a; Hamilton 1964b; Hamilton 1967), assumes that the objects of selection in evolution are not individuals, and certainly not populations or species, but only the alleles of the individual genes. From the viewpoint of this hypothesis, all genes are selfish or, to be more exact, all the alleles of all genes are selfish. Selfishness here means that every allele is “out for itself”. Only an allele that affects the properties of an organism so that it increases the probability that it will be replicated and transferred down to future generations more frequently that other alleles of the same gene can be successful in evolution. In most cases, a successful allele somehow increases the biological fitness of the individual in whose gene it is contained. Consequently, the selfishness of genes is not fully apparent at first glance. It might seem that alleles that bring an advantage to individuals are most successful in evolution.

An ultraselfish gene is a gene or, to be more exact, an allele that, in order to increase the probability of its proliferation in the gene pool of the species, reduces the biological fitness of its bearers (see, for example, the bluebeard model in Section IV.9.1).

Molecular drive is a process through which mutations can proliferate within gene families (in process of homogenetization) and within the population (in process of fixation of mutations) through a number of mechanisms of nonreciprocal transfer of genetic information occurring on the chromosome or between different chromosomes (Dover 1986). Molecular drive differs from genetic drift in that changes in the frequencies of the individual alleles that occur through its action are not random in their direction. If a certain population of genetically identical organisms is divided into several smaller populations, then genetic drift will lead to fixation of different alleles in each population. In contrast, the effect of molecular drive should lead to fixation of the same alleles in all populations. Molecular drive differs from selection in that the alleles that are fixed through its action need not favourably affect the phenotype of the organism and can thus have a zero or even negative impact on the biological fitness of the individual. 

In molecular drive, one allele is replaced by another not because this is more advantageous for its bearer, but because, at the level of the DNA, it multiplies more effectively, either through a mechanism related to replication or through a mechanism related to gene conversion (see below).

Molecular drive differs from mutation bias and reparation drive mainly in that it is responsible for the proliferation of certain mutations in the genome or in the gene pool of the population, but not for their repeated formation.