Scientific collaboration is a challenging proposition, most scientists would agree that collaboration is a key component in problem solving, but few practice it beyond their immediate research group, as their careers are dependent on publishing. As a result, they hoard ideas, data, methods, and all the other stuff necessary for progress, and publish it themselves.
The culture of the scientific community is geared to recognize publishing new and original stuff in scientific journals, not sharing ideas on a wiki. Scientists do not get a job or more funding on the basis of wiki-sharing great ideas, but they do for getting marginal stuff published, so guess where the focus is! To build collaborative scientific effort, we need to reverse this relationship.
During the project to map the human genome, huge amounts of data were required, data that was dispersed amongst the various bodies doing the research that generated it. To facilitate sharing, an agreement that became known as the “Bermuda Principles” was forged that created the culture of sharing data immediately it became available, and this simple change turbo-charged the effort to complete the project.
What drove the difference the agreement created to most other scientific collaborative efforts was that the major funders agreed that the price of participation in the project was the sharing of data, if you wanted the funding, the absolute condition was sharing data. Bingo, collaboration was created by putting a price on participation.