I was performing various ligand docking procedures using Rosie and when I was analyzing the data I noticed that some decoys had multiple entries in the score file with different scores but the same decoy model name. Is this normal and if so how can I interpret it?
For example one of my procedures created 200 docking models but the score file had 204 entries, 4 of which referenced already referenced models.
In general, the numbering of the output structures is arbitrary. Depending on how things are run, you can possibly get re-runs, where Rosetta doesn't realize it already output a structure 143, and then re-runs it. When it re-runs it, the trajectory will be a completely fresh one, so you wouldn't expect it to recapitulate exactly the same energy/structure.
So I wouldn't be too worried about the multiple lines in the score file, even if they have different energies -- they're just re-runs, equivalent to . The trick would be in matching up the score line with the structure, in the event that that particular output is one you wanted to look at further. But there should be scoring information included in the output PDB file which should allow you to figure out which "143" structure you have. (Unfortunately, the structures for the other score line will be lost, either as a cause of or a result of re-running things.)
Sounds good, thank you!