There are 5-10 small to medium-sized companies active in this space along with various academic efforts. The biggest problem for low or no cost alternatives has been the difficulty in creating a graphical front end of merit, or agreeing which promising candidate will be generally supported. To date, it’s been a small market with huge upside that’s not been realized (drugs can make 1-2B per qtr).

The problem with data mining (which has been done in various forms since the 50’s) is the quality of the data and it’s (in)consistency.