Comment by macleginn
7 hours ago
The selection is classical, but the mutations are supposed to be random, not skewed towards more useful genetic variants. One can speculate that a purely random (and otherwise neutral, say) prior mutation led to this directionality, but the actual resulting mechanism is AFAIK still unknown.
Natural selection has nothing to do with randomness per se, it’s merely differential selection based on environmental fitness. The fitness function of a given environment can even narrow a range of existing genes without any mutation occurring during the period where the new fitness function is applied. The classic example of this is peppered moths shifting toward the darker range of their preexisting color range during the Industrial Revolution.
Horizontal gene transfer is another way organisms can acquire new traits.
plasmids
https://en.wikipedia.org/wiki/Plasmid