Between 80% and 98% of our genome is junk DNA - meaning that no function has been identified for it (yet). Does that mean that this DNA really does not have any function (anymore), or just that we haven't figured it out ?
Few people outside the field of genetics know that our genome is affected by viruses, who have through hundreds of millions of years of evolution, slowly infiltrated our DNA sequence.
A virus has the size of a small gene. Although it is not clear yet how it happens, a virus can place its genetic sequence inside our own, and stay there for subsequent generations, pretty much forever. In most cases it does not have any effect on the host, but occasionally a virus can intercalate itself inside a working gene and disable it or modify its function. This is one cause of evolution (the other being spontaneous mutations from generation to generation). In most cases it will kill the host, or leave it with a disability. But sometimes it will improve its health or confer it some new attribute, which will give it an evolutionary advantage.
Junk DNA also consists of a large number of repeated copies of a short DNA or RNA sequence, which are apparently transcription mistakes that have accumulated over the generations. It has been estimated that 42% of the human genome is formed by such retrotransposons, as they are known, including 8% caused by Human Endogenous Retroviruses (HERVs). The latter are viruses that duplicate their genetic code within our genome just as a way of propagating themselves and ensuring their own survival. They are selfish genes par excellence, just caring about their own survivals, even if it is detrimental to the host.
A small percentage of retrotransposons are copies of working DNA. The number of repeat copies of a DNA sequence within a gene can vary from person to person and has a direct effect on our health or behaviour. For instance, people with 7 or more copies of a specific 48-base pair sequence in the dopamine receptor D4 gene are more novelty-seeking and at higher susceptibility of developing ADHD than people with 2 to 6 copies. This is because having more copies increases the dopamine in one part of the brain.
A third category of junk DNA is the disabled genes. We carry with us a lot of information about our ancestors. As apes, we don't have a tail anymore like other mammals or reptiles, but the gene for making a tail is still within our genome. The gene has just been deactivated. Theoretically, and ethical considerations aside, it would be possible to create humans with a tail by reactivating the gene by genetic engineering. There has been a lot of evolutionary changes from the bacteria to us, and therefore a lot of our DNA is also composed of old relics of this past.
What I am wondering is if this Junk DNA is really all useless (or even detrimental), or if it is necessary for our proper genetic functioning. Scientists have tried deleting some junk DNA from mice, and they seemed to be just fine.
But what would happen if all the junk clustering our genetic hard disk were to be cleaned up ? some working genes that had been separated by junk buffers might recombine and work differently, or stop working. For all we know, our body might need these retroviruses to fight other viruses, or prevent new insertions in "strategic parts" of our genome in future generations. The more junk there is, the less likely it will be that new HERV's will intercalate them inside working genes. If all our DNA is useful, then the slightest copying error or retrovirus contamination would have devastating effects on our offspring. So we might need that Junk DNA after all. On the other hand, it might also be possible to improve our health by removing a few noxious elements from our genome. Most important of all, some of the so-called Junk DNA might have vital functions that has not been identified yet.