This is probably quite a basic question, but I found no reference to it.....

1. For the results we return, what is the difference between a "starting structure" [always just one returned], and a "decoy" [multiple ones returned per run]?

2. Is the starting structure a basically stable structure, and then the decoys are results of "fine-tuning" the starting structure to get an even lower energy?

3. And lets say there was no limit on run time, would there be an almost endless number of decoys for each starting structure?

4. And why not allow run times of longer than a day, in order to generate the maximum number of decoys?

Thanks!

---

Regards,
Bob P.

I'm not an expert but here is what I think. A starting structure is the state of the molecule before you start optimizing, a decoy is the result of the optimization (lowest energy structure).

If there was no limit on run time there would be virtually endless decoys. And one of them would be the actual folding of the molecule.

There is virtually no maximum of decoys.

Many computers generate decoys, they are sent in and the best one is picked. It would take years(or more) to do that with one normal computer.

I hope this anwsers your question and that I didn't make any horrible mistakes.

https://boinc.bakerlab.org/rosetta/forum_thread.php?id=4109&nowrap=true#53399

mod sense explains it here

There are actually so many possible conformations of the atoms in a protein that it would actually take billions of years of computing time to consider each one. Rosetta is designed to find the conformation without having to actually run through every single one of the possible orientations. The idea of course being to find ways to accurately skip only those orientations that are not the native structure. The more you accurately skip, the faster you arrive at your solution. On the other hand, if you skip too much, you might miss the most accurate solution. The trick is in knowing the difference.

---

Rosetta Moderator: Mod.Sense