On patent applications

In the last issue of Nature Biotech there was a list of recent patent applications in systems biology. Most of them are general ideas that have been published several times with different methods. I wonder how can a patent like this be approved and enforced when there are so many research papers available on the subject.

An example:
"A tool providing interactive capabilities for user involvement in extracting and disambiguating biological information in text; useful in generating a biological diagram."
A simple search for text mining in Pubmed will point you to more than 100 papers for example.

The same holds true for most of the other applications you can find in the table.