FINALLY, A COMPUTER I CAN LOVE

Out to dinner.  Where?

Many years ago I was an avid reader of science fiction.  I liked it so much that, while supposedly working

on my M.S. degree in geophysics, I took time out to submit a story to Amazing Stories, which at the time was the premier magazine of that type in the world.  In due course I received a polite rejection from the head editor, who encouraged me to try again.  Real life intervened and I never did.  That’s probably for the better, but who knows?

Anyway, I vividly remember a short story from those days, in which a computer was being awarded the Nobel Prize in medicine – for curing cancer.  It made an acceptance speech, if I recall correctly.  Did it wear a tux?  Or an evening gown?  Can’t remember.

So, maybe that wasn’t such a far-fetched concept.  Dick Ingwall, trapped in his house like me by gale force winds ripping through the Borrego Valley, has studied the NY Times and come up with the following article, which I will comment upon shortly.

http://www.nytimes.com/2014/03/27/science/enlisting-a-computer-to-battle-cancers-one-by-one.html

The deal is, sequencing DNA has become so easy, rapid and cheap that it is possible to take a sample of a tumor and quickly and economically determine which of its genes are mutated.  This doesn’t get you very far, however.  Some of these mutations will turn out to be harmless and should be ignored.  Others are probably harmful, but it isn’t always apparent how they harm, nor what you can do to stop them.  Here is where Watson comes in.  Watson is the IBM super-computer that apparently just starred on Jeopardy.  Watson is busily devouring the abstracts of medical articles – it is up to 23 million or so.  Give Watson a list of mutations and it will spit out a list of drugs, or drug cocktails  that might help.  Then, of course, humans enter the picture.

Obviously, Watson can only make suggestions based on the reported research.  If the research is crap, Watson will mislead.  That’s hardly Watson’s fault.  The familiar adage “garbage in, garbage out” applies, even to a super-computer.  We need lots of basic research for Watson to devour.  Lots of good basic research.

This approach to cancer therapy is going to be tested on a group of 20 victims of a nasty form of brain cancer, glioblastoma.  Prognosis from glioblastoma is dismal.  Let’s hope Watson gets it right.