- This topic has 4 replies, 3 voices, and was last updated 11 years, 10 months ago by Anonymous.
January 15, 2012 at 6:00 pm #1138Anonymous
I’ve searched for details of what exactly the relax protocol does. I know that it does “agressive sampling” and basically just minimises the Rosetta energy. Is there anywhere I can get more detail than that? I’ve looked through the Rosetta publications and many describe different aspects of Rosetta but I haven’t found any that describe specific protocols or applications. I understand that these protocols keep evolving and therefore cannot be described “for once and for all” so, is my search in vain?
I’m writing my thesis and I used “relax” to refine models. I would like to give some background info as to what it actually does. I chose the relax protocol based on the descriptions of the different protocols in the manual and after reading many favourable publications on Rosetta itself.
Any reply or just a link will be apreciated!
January 15, 2012 at 10:42 pm #6520Anonymous
To get the level of detail you want, you’ll need to either read the code directly, or talk to the folks that wrote it. You may be able to find detail at this level in theses of the paper authors (those who were students for the papers) if you need something citable. I’ve emailed a couple of relax-experienced folks to let them weigh in.
January 31, 2012 at 1:46 pm #6598Anonymous
Hi, thanks for your replies.
I’ll read the paper mentioned and see if that helps. I think I’ll be able to get away without too much detail but was quite interested myself as to what it actually does.
I guess only when the protein folding problem is solved and the software stops changing will it be worth properly documenting everything to the public.
January 15, 2012 at 11:00 pm #6521Anonymous
Mike Tyka in the Baker lab is the person who has done the most work with the relax code – looking at publications by MD Tyka is a good place to start. The recent “Algorithm discovery by protein folding game players.” paper by Khatib et al. in PNAS will likely be a good read – the algorithm it talks about is a variant of the relax protocol, and has some discussion of the comparison.
Keep in mind that there are several different relax protocols, each accessible through the relax application. I’m not sure about 3.1, but I don’t believe the default for 3.3 (“fast relax”) has an official reference paper yet. (Although it’s discussed in broad brush strokes by Khatib et al.)
January 31, 2012 at 7:09 pm #6600Anonymous
The lack of good documentation on a large number of protocols is something that frustrates even Rosetta developers. We’re trying to get better on putting out good documentation (e.g. the RosettaCon2011 Tutorials Collection), but we have a ways to go. Hopefully we won’t have to wait until the protein folding problem is solved before we get decent documentation, but you’re correct that continual improvements complicate producing it.
That said, it’s worth pointing out that you have access to the code. I realize that not everyone can read C++, and Rosetta code tends toward the complex/opaque ends of things to start with, but the code is really the best place to look if you want to know nuts-and-bolts details of how things are done.
- You must be logged in to reply to this topic.