I would like to add in regards to the Remix tool that I have never been able to find a Remix replacement when an end segment is included in the selection. It may be just me but I would like to have an explanation for this. And if a solution is not possible with Remix, what is proposed to enable us to script rebuild-like moves to end segments (other than with bands)?
FoldIt seems to have got stuck in a bit of a cycle recently. The top scoring monomer design puzzles (usually a linear-looking structure with 2 helices and 3 sheets) get recycled back as de novo puzzles: I've also seen these structures in Rosetta@home. These structures are clearly interesting but it would be good to know why.
Could someone comment on the internal code:
LocalWiggleSequence in GUI versus structure.LocalWiggleSelected() in LUA:
Are they the same internally?
Or what would be the equivalent LUA function for this GUI one?
Extract from chat on 2017-01-13:
22:09 Wbertro TomTaylor5: converted "Wiggle by sheets" to LUA
22:09 TomTaylor5 Great
22:10 Wbertro but the code does not give the same results as the GUI one
22:10 Wbertro however it does not crash the client at the end
22:11 TomTaylor5 What would you like to hear first? The good news or the bad news?
22:11 Wbertro I think the GUI LocalWiggleSequence code is NOT the same as the LUA LocalWiggle code
22:11 TomTaylor5 That was probably the function I couldn't find.
22:12 Wbertro but the 1322 puzzle I tested them against is so sensitive that two runs don't give the same result twice
22:12 Wbertro I don't think it is a missing function
22:13 Wbertro it is simply DIFFERENT internal code, it seems
22:13 Wbertro I think I will ask on the next chat
Previously unanswered chat questions:
http://fold.it/portal/node/2002443#comment-32490
Apparently, these questions have been asked at least a few times before, but have either never had any clear answers from the Foldit devs, or need to be revisited.
Q1: Will a parallel programming language, such as CUDA or OpenCL, ever be used to optimise processing speed in Foldit?
Q2: Will Foldit ever be open source?
Currently, only Rosetta is open source, not Foldit.
https://developer.nvidia.com/cuda-zone
https://www.khronos.org/opencl/
Previous parallel programming discussions:
https://fold.it/portal/node/988293
https://fold.it/portal/node/991445
https://fold.it/portal/node/992230
https://fold.it/portal/node/1998564
https://fold.it/portal/node/1998734#comment-29310
https://fold.it/portal/node/1998822
https://fold.it/portal/node/1998865#comment-29346
Previous open source discussions:
https://fold.it/portal/node/164128
https://fold.it/portal/node/986241
http://fold.it/portal/node/986352
https://fold.it/portal/node/99390
When ring sidechains are near each other, does the scoring algorithm reward them just being close together, or does it distinguish between good stacks and bad stacks? I know they aren't always face-to-face in nature; sometimes they are perpendicular, or parallel but offset. But the foldit score seems to reward proximity even when their angle relative to each other seems pretty much random, or one is bent away from the other.
When you have a foldit design that looks promising, but Rosetta doesn't settle on a single fold, do you try different sequences on that backbone to see if one of them gives better Rosetta results? Lin & Koga et al designed their backbones first and then used Rosetta to figure out what sequence would make that backbone work. Seems like the same could be done with promising foldit designs. We players are running a very small number of clients, so we can only pursue one or two or three promising sequences - using Rosetta@home would make it possible for you to test many more sequences on a given backbone.
Sometimes foldit picks an unlikely AA (like a serine in the interior just because it is crowded there) that might make the fold fail in Rosetta and in real life, but if that AA could be fixed then the fold might work.
In figure S7 of the Lin & Koga 2015 ideal design paper (supporting information), they show which ideal loop combinations Rosetta liked best. For helix-sheet loops, BAAB (blue-red-red-blue) worked more often than GB (green-blue) or GBA (green-blue-red), but we do not have BAAB in our building blocks. For sheet-helix loops, BAB (blue-red-blue) competed well with GBB (green-blue-blue), but we do not have BAB. Would you consider adding the missing loops to the building blocks?
Is there a plan to expand the Blue Print and remix databases?
This question is one I would have posted. Perhaps I may go further…
I can understand the logic for these tools/filters….seems sensible to me. But I also believe the existing databases
to be small at this time (since the concepts are effectively in a testing phase perhaps)
But here's my question….
At the moment it feels as tho the libraries are small for both the reason above - and Lab preference.
Should these tools continue (and I hope they do) will the libraries have an emphasis on what the lab would like
to see - rather than what real life exhibits? Or will they be more fully populated without heavy, directional intervention
on the part of FC?
I finally got a version of two hot dogs surfing past all filters. I shared my first version with the scientists as I once again "passed" the filter even with two backbone problems. It is the one with the score below 4000.
I know Gitwut also shared some funky examples, and I have shared one with one backbone problem before. Bottom line, I don't know that we have a big enough library or the tools to work with these puzzles and produce what you want. For example, maybe we need to rebuild two loops at the same time to get things fixed…. maybe "ideal" needs a bigger library and some tests
Direction would be helpful. A Koga style fold will not fix to ideal unless I figure out how to rework two loops at the same time – and I have no idea now to get Blueprint to do such a layout so I am still hand folding to start. We are headed the right way - we just need more feedback and help
One suggestion – perhaps we can get a share with programming track in addition to share with scientists. That might help us separate process from content. But I'd still like help with both
Skippy