This is a subtle one. The headline (New Scientist, 2 May, p. 11):
Gene discovery
may be common
cause of autism
My first reading of the compound noun gene discovery is that discovery is an abstract noun, referring to an event (in which some gene, or possibly genes, is discovered, though in other cases N + discovery could refer to an event of discovered by N(s), as in a University of Chicago discovery; there are both “object” and “subject” readings of compounds with abstract nominal heads).
But it’s ridiculous to asset that an even of discovery is the cause of any condition. Something like “The discovery of genes at the University of Chicago may be a common cause of autism” is, at least at first, puzzling. (Ok, here’s a science-fiction scenario to write about.) Instead, N + discovery is intended to refer to the thing discovered about (in this case), or by (in other cases) N. This reading is available in the New Scientist headline, but it takes a little work to get to it.
You see the headline writer’s dilemma. The writer was given a very small space to produce a head, and the obvious “Recently/Newly discovered gene may be common cause of autism” won’t fit. “Newly found gene” might have fit, but give the headline writer a break.