Hi, James

There is at least one level of protection. We are accumulating a number of 'results' from each Work Unit. If you are looking at the Application View you can see progress measured to 1 decimal place. When a 'result' is completed, the progress will jump up several tenths of a per cent. The result is encrypted and added to a temporary checkpoint file on your hard disk in the WorldCommunityGrid directory. Which is why the hard disk is unlikely to power down. When the program completes its assigned number of 'results' (in a non-deterministic amount of time) the checkpoint file is read and decrypted. Then the entire file is compressed and re-encrypted. Then it is sent to the server. Once the server acknowledges receipt, the points awarded are recorded in the Agent but will not show up on the batch-processed Statistics pages until the next day or the day after that.

This same project is also running on GRID.ORG ( http://www.grid.org ) and reading the posts on their site I have learned that their server is using a "What I tell you three times is true" algorithm. Three different computers run the same work unit. If they do not return identical results then presumably more run it. The idea is to catch errors from spurious signals / faulty motherboards but it also offers a level of protection against malicious PC contributors. I presume that the World Community Grid does the same.

As for the importance of the project - - the molecular biologists will have to answer that. Back in school I was taught that progress is made where it is made and that scientists have a hard time knowing in advance what is vital and what is trivial. Still, now that the Human Genome Project has mapped out the protein template regions on the DNA strands, an obvious Step Two is to determine just what molecules are actually produced from them. The long strand of protein bases crumples up (folds up) and forms all sorts of interconnections. The result is a huge interconnected molecule. This is obviously a tremendous combinatorial problem. Knowing the DNA template makes a quantitative analysis trivial, but the qualitative analysis is extremely difficult. Just how is each molecule folded up?

And of course, a number of proteins undergo significant processing and no longer directly match the DNA template. But that is Step Three and I think that we are concentrating on Step Two right now. [Grin]

Lawrence