Would you like an unexpected recombinant protein with that, sir?

Read about this  little oops.  Untested genetic code variations going out into the world.  This is subject of interest to me.  My original undergraduate degree was in Genetics, so this at one time was a subject that I knew a bit about.   

The whole genetic engineering industry is based on the idea that if you take a gene out of one context, it will perform the same function in the new environment, i.e. from bacteria to corn, one should get the same protein produced.  This assumes that the DNA gets read, gets transcribed into RNA, which then goes and links up all the amino acids that you need to make a protein.  Now at a basic  level this seems like an interesting idea.   Same basic process, same results.  This is called the "One Gene, One Protein" idea.  Only problem is that it doesn't work. 

I am adding an article here  from the 2002 edition of Harper's, which is as good a basic introduction to the realities of the situation as I have seen.  Basically the proteins that are strung off the RNA that came from the DNA work because of how they fold up.  And they get all sorts of other elements added to them, like sugars and so on.  So basically, the raw string of material that came off the DNA is likely OK, but the post processing environment is totally different.  To give an analogy, its like sending car parts to an aircraft manufacturer, and then wondering how come it is that they don't do the job. 

What we have here is a lot of room for unexpected consequences.  Now, I like complex systems, and I find their various failure modes a fascinating subject to analyse and discuss.  What I have to say about this is that the system that is being stress tested here is the same one that I am relying on to run my consciousness, and that's a pretty big failure if it goes wrong.