The sickle-cell story is a little more complex than we're talking here.
Every human has two copies of every gene (we are 'diploid' organisms). If a person carries one copy of the sickle-cell mutation then they are resistant to Malaria as described. However, carrying two copies of the sickle cell mutation is seriously detremental and potentially fatal.
So there are two forces in the system : malaira (environmental factor) keeps the sickle cell gene around because having one copy is beneficial, but full-blown sickle cell (genetic) puts the brakes on because when two sickle cell carrying, malaria-resistant parents have a child there is always a 1 in 4 chance of that child having full-bown sickle cell and never reaching adulthood.
This is why not everyone has sickle cell.
Biology is a constant system of balance and tipping points.
If you really want to know how DNA turns into organisms, you'll need to read up on developmental genetics (a vast and complicated subject). The Homeobox genes are a good place to start.
Homeobox is the gene that tells the fertilised egg which end is the head, which end is the tail and where everything else in-between should go. In the fruit fly (Drosophila) , the egg receives a dose of homeobox protein at one end while being laid by the mother. This dose diffuses through the length of the egg forming a gradient of concentration. The concentration of homeobox protein is a trigger to different genes at different homeobox protein levels; head genes are switched on at one end, tail genes are switched on at the other end and everything else is switched on in the appropriate place.
The homeobox gradient is actually just the starting point for a whole cascade of protein gradients that map out the body plan in the embryo. The whole thing is an amazing analogue control system (sorry geeks, we’re analogue not digital).
These gradients sometime cause strange mutations : Anne Bolyn was said to have six fingers (which is a known mutation caused by a particularly strong concentration of a developmental protein in the area of the thumb (the concentration of the gradient determines the positions of the fingers, if it’s too strong fingers continue to grow until the concentration drops away). I once read a wonderfully titled paper : "Why Great Danes sometines have an extra digit, but poodles never do" which covered this subject. Sadly I can't find any references to it on-line.
I think all software developers should study the cascade system of homeobox genes : if nothing else it will stop you using cascades of events in your code : it’s just too complicated!
Herbie