pesto writes:
In this case we could use vaccines instead of relying on this potentially harmful gene.
There are vaccines, but its hard to get them to the vast population that needs them. The latest approach is to genetically engineer mosquitos that are resistant to the malarial parasite. If we can get a gene for malarial resistance to go to fixation in the vector population, that might be a semi-permanent solution, or certainly the most cost-effective one, while it lasts.
pesto writes:
Which would be worse, having a population at risk of Sickle Cell disease or at risk of Malaria?
Yes, and I think there are better ways to deal with malaria than counting on the sicle cell gene. But you see how this gene has evolved to be as prevalent as it is? It is only deleterious in the homozygote. So let's say the frequency of the SC allele is 1/20 in a population. The people actually suffering from SC is only 1/400 (we square the allelic frequency), whereas twice as many people (1/200) have the benefit of being a heterozygote.
pesto writes:
So, in this case, could we safely do away with the gene that causes haemophilia?
I expect so, and another might be the one that causes phenylketoneuria, the inability to metabolize phenylalanine. But you can never really 'do way' with them completely because they each probably have a 'de novo' mutation rate. They'll be recreated again at some point in a germ cell division and you'll never know where or in whom.
pesto writes:
So, a bigger question might be, how would we decide which genes we could safely eliminate?
Exactly. The 'black and white' mutations that are obviously deleterious are in the minority. The bulk of mutations are in some gray area and we have no way of judging their potential 'value' outside of present environmental conditions.