Selfish DNA is non-transcribed, non-coding, and contributes nothing to the well-being of the organism; in most cases it is selectively neutral.
Once it arises, selfish DNA is passively replicated and passed on from parent to offspring. Changes in its frequency in the population are due to drift. Only if a non-coding sequence interfered with the construction of the organism or if it accumulated to such an extent that the cell cycle was slowed down by the need to replicate it all, would selection act to reduce it.
Provided the quantity of a particular sequence is not excessive, it is not transcribed, and it accumulates in parts of the genome where is does not interfere with genetic regulation and transcription, there is no reason why selfish DNA should not evolve.
There are two types of selfish DNA:
• Passive DNA. The sequence itself might not influence the chance that it spreads in the DNA and is retained; it is then a passive kind of selfish DNA. It could accumulate as 'junk DNA' in the genome.
• Parasitic DNA. A particular sequence might have a better than average chance of spreading through the DNA; these sequences would be a more active, parasitic kind of selfish DNA and would proliferate until checked by natural selection.
The image opposite is of the toad Bufo bufo:: only 20% of its genome is single copy DNA. If only the single copy DNA is the coding part, then 80% of this toad's genome is non-informational. Selfish DNA can explain this.
How might selfish DNA originate?