It is frequently difficult to draw a sharp boundary line between mutation bias and reparation drive. Both processes have very similar external manifestations and similar or even identical molecular processes are responsible for them both. Nonetheless, it is apparently advantageous to differentiate between these closely related processes. While only the chemical-physical properties of nucleic acids or molecules that interact with the nucleic acids are responsible for mutation bias, reparation drive is a process whose manifestations are, at the very least, partly tuned, i.e. from a functional standpoint more or less optimized through natural selection. As a consequence of this tuning, mutation bias and reparation drive frequently act in opposite directions and neutralize one another in their effects. For example, methylated dinucleotides CG frequently mutate to TG, as deamination of methylated cytosine yields thymine. Consequently, the cell nucleus contains molecular repair systems that preferentially replace nucleotide T by nucleotide C at sites where the G-T pair is present instead of the G-C pair in opposing positions on the DNA (Brown & Jiricny 1988). Where the mutation actually occurred through deamination of methylated cytosine, the repair system renews the original DNA sequence (Fig. VI.1a). Where the mutation occurred through some other mechanism and, on the other hand, the incorrect nucleotide is G in the opposing strand, the repair mechanism conserves the mutation (Fig. VI.1b). In DNA segments where deamination occurs very frequently, i.e. mutation bias leads to replacement of nucleotide C by nucleotide T, in segments where these mutations do not occur, in contrast, nucleotide T is replaced by nucleotide C through reparation drive.