'course, this all assumes a definition of "information". Is there a generally accepted definition of "information" in the context of a genome? I suspect not.
Dembski has been doing some work in this field. You might find this interesting: http://www.designinference.com/documents/2004.08.Variational_Information.pdf