to those who have figured this out?
Wouldn't they be able to load the native structure from a protein database and use that as a guide toward folding the puzzle proteins, simply trying to make their own solution match the one from the database?
Can they identify an anonymous protein by using the sequence of amino acids supplied with the puzzles to look up matches in a protein database and find the actual protein being folded?
You also found out how to load pdbs :), and i agree its a pain. I would like to use fold.it as a pdb viewer as well. It's one thing to see them in rasmol or another viewer, but to "feel them" and play with them is just fun.
I think this issue has been raised before. I don't believe they wanted us to see actual proteins, or just look up the one we are working on as they are interested in how humans fold proteins and overcome problems, not so much how we can copy a pdb.
Then there is also the issue of what the client can support, which so far, seems to be only amino acids. not any other atoms you may find in a pdb, causing it to crash out. You may have seen it if you ever tried to load a pdb with some HETATMs. I also couldn't get it to read structures… so they get imported as all loop.
You also found out how to load pdbs :), and i agree its a pain. I would like to use fold.it as a pdb viewer as well. It's one thing to see them in rasmol or another viewer, but to "feel them" and play with them is just fun.
I think this issue has been raised before. I don't believe they wanted us to see actual proteins, or just look up the one we are working on as they are interested in how humans fold proteins and overcome problems, not so much how we can copy a pdb.
Then there is also the issue of what the client can support, which so far, seems to be only amino acids. not any other atoms you may find in a pdb, causing it to crash out. You may have seen it if you ever tried to load a pdb with some HETATMs. I also couldn't get it to read structures… so they get imported as all loop.
Yes Brick, that's the problem. Anybody, with a little work, can track down the actual PDB used. It's possible to identify the protein using just the AA sequence. So I guess the question is, what the dev team thinks of this? Should we ever know the PDB ID?, see a native structures?, find similar proteins to the one we are working on?
If the PDB IDs are known to everyone then it wouldn't be as unfair, but so far I've only seen one recent puzzle where the ID was mentioned right in the description. 124: PDZ Domain Protein "…the initial model for this puzzle is derived from the PDB code 2fne"
I do believe being able to load the native form of an active puzzle in the fold.it client may be a bit to much of an advantage… if we are ever allowed to load a .pdb into the game, maybe it could refuse to load any that have 90% or more of its AA sequence in an active puzzle.
The client should not be able to load PDBs. We were aware of that you could due to an oversight, but we fixed that 2-3 updates ago. We don't plan to make Foldit a general PDB viewer at this time.