If you can get a Goodman and Gilmans reference book you'll get more info
than you seek.

You seem to be laboring under the asssumption that the retail price of a
drug should be connected to the manufacturing cost.

The cost of selling a drug comes long before manufacturing.  The company
must discover a potentailly important chemical, then through animal
testing prove that it is unlikely to cause harm if administered to
humans.  The company can then carry out the testing on humans to
demonstrate that the drug is both safe and effective.

8-12 years and $500 million to $1 billion later, the company can gain
approval to sell the drug.  In selling the drug, the company must cover
the cost of developing it, plus the cost of developing the ones that
din't work before the patents run out and anyone can sell it.

If you want a comparable, consider the software industry.  How much does
it cost to burn and ship a CD.  And if the software is delivered over the
internet, the manufacturing cost is effeectively $0.  Copying software is
certainly not difficult or labor intensive, so why is it so expensive?  
Infact, on a weight basis, software is far more expensive than many
drugs.

If you are interested in what it takes to bring a drug to market, take a
look at the PhaRMA website.

Some drugs are semi-synthetic; cafergot, an ergot alkaloid, is a
semi-synthetic.  Most, if not all, SSRIs are synthetic--very simple
chemistry.  The goal of the industrial chemist is to obtain a very
pure product with as few steps as possible.  No exotic reagents and no
lengthy steps in purification--if it can't be scaled up to plant
production, it isn't worth it.  If a bug can ferment something and
yield a precursor, start there.  The closer to the target, the better.

Marketing and trials are where the big bucks are spent.