Feature Request: Walk the chain maga-command

Case number:845813-730864
Topic:Game: Tools
Opened by:jas0501
Status:Closed
Type:Suggestion
Opened on:Thursday, December 18, 2008 - 09:32
Last modified:Wednesday, June 29, 2011 - 18:36

I think it would be very useful to provide a mega-command along the following lines starting from an unlock position

a. Set a lock at position n
b. set a lock at position n+k, where K> 1 and is small
c. local wiggle between locks
d. stop wiggle
Walk the chain via
--------------------
e. unset lock at n
f. set lock at n+1
g. unset lock at n+k
h. set lock at n+k+1
--------------------
i. goto c

Note: step d. stop wiggle would be a user supplied event.

This would be useful for walking the entire chain as well was walking hot spot sections.

Assumming only 2 locks are set al that is required is a direction to advance.

The above description does not address the tail ends, that is left as an manual exercise for the user.

Comments anyone?

(Thu, 12/18/2008 - 09:32  |  7 comments)


Joined: 10/19/2008
Groups: Oma Gawd

"Proper" isolated wiggling, ie walking the backbone, is not as simple a procedure as you describe. And establishing a narrow, simplistic interpretation of how it is done is, imo, detrimental to the open use of the procedure by ALL users. One must carefully evaluate a "best" procedure long before implementing such a sweeping change or face the high likelihood of devolving the game into a simple subset of rosetta, ie, a mindless, plug-n-chug solution, rather than a human-intuition powered game resulting in "quantum" changes in score ie folding, (which is, according to the game definition, what we're supposed to be providing). That is, mega commands like this imo lead to a grunting brute force solution, where the only issue is computing power, rather than mind power.

Joined: 12/08/2008
Groups: None

xaiando said "mega commands like this imo lead to a grunting brute force solution, where the only issue is computing power, rather than mind power"

--------------

Well, I partially agree, however the mega-command can be targeted at a small portion of the entire length and is not just a brute force implement.

The advantage of this type of command is also instructional assuming the involer is paying attention. Observing the areas that are hot during the walk can be very enlightening.

Obsering the behavior of varying lengths being wiggled can also be very instructional.

The insights gained observing the walk may translate into an appreciation otherwise not as easily gained.

I think avoiding some of the drudgery of repeated walks over the same portions time after time for the slow tightening of the protein can be alleviated.

-------------

One could argue that the shake command is a brute force mega-command. One might gain some valuable insight doing manual shakes but the effort is not worth the gain.

-------------

As the "players" gain skill and the protiens get more complicated I think the addition of user constructed "macro" commands with save time and preserve sanity and improve the quality of the results.

-------------

One can use a hand saw or a power saw. In the end the house gets built either way. I like my power tools.

Joined: 10/19/2008
Groups: Oma Gawd

is that your method is essentially "fixed".

without providing a tacit programming system within the game, it means that any such addition requires using one person's method over another. I personally disagree with your paradigm for automated wiggle. I think it is rough-shod, and I would never do it in the manner in which you describe.

the problem with providing a programming environment is that in doing so, we stand a greater chance, in fact a very high likelihood of making this into little more than Rosetta. While Rosetta is a fine project, the "participant" in Rosetta simply watches the machine go. It is a "game" of who has the largest computer farm.And that is, to the best of my knowledge directly opposed to the fundamental premise of this project, ie game. Within a programming environment I could do everything I am doing now, just automated..how freakin boring. Yeah...lot's more kudos points for the BOINC crowd as they toss their computer farms at it...which is even more boring.

this is what I mean by grunting, automated solution. I realize the value of automation, having produced many valuable hardware/software systems towards that purpose for the aeronautics, research, medical, and industrial communities.. but the fundamental premise given to the participant at introduction to this effort is that it exploits mind power from a wide variety of people, without any need for scientific background (of which programming is a subset), not just computational power and programming skill.

As I said, I do understand the obvious value of automation...However, the recent update that broke shake and took isolated wiggling from a secondary or even tertiary role to primary in score accumulation is, imo, giving a false impression of it's actual value, as I believe that the bulk of those points are only available via isolated wiggle *because shake is broken in the first place. Unless of course, the developer's paradigm has shifted to automation of this project...(I have had no indication that this is so)

My understanding is that any algorithmic improvements gained thru study of foldit participant's behavior would go to their other project, ie, Rosetta, not duplicate it's purpose here.

admin's picture
User offline. Last seen 5 weeks 2 days ago. Offline
Joined: 11/10/2007
Groups: vi users

I'm changing the priority of this. Forgive me if you disagree with this, but it is a large undertaking and cannot really be called urgent.

Joined: 12/08/2008
Groups: None

That is fine.

I must admit that as a newbie my perspective is a bit naive regarding "optimal" approaches to folding. I think my post was a bit misinterpreted in ant I am not suggesting this commnd, or camoonad like this, be started and you go have lunch while it traversse the hold chain. I aggree this is counter to the intent of FoldIt/

I did indicate that a manual stop for each wiggle is required. I do think that some macro cammand or command composer mechanism would not be counter to the program goals. The key point is that the user still has to indicate a stop of the interation or od the current sub command.

An iteresting question is whether providing some macro language for users to compose an exection outline and invoke it is counter to the Foldit Philosopy.

As an example if I know there are 3 hot spots and I want to wiggle each, is preparing a "script" named "3 spot wiggle" of:

clear all locks
lock 27
lock 30
wiggle 28 and wait for user stop
reset to max
unlock 27
lock 28
unlock 30
lock 31
wiggle 29 and wait for user stop
reset to max
lock 45
lock 48
wiggle 46 and wait for user stop
reset to max
lock 54
lock 57
wiggle 56 and wait for user stop
reset to max
wiggle and wait for user stop
shake

and kicking it off really any different than manually doing the same thing? As a game player, if I'm going to do this exact sequence 3 or 4 times, is kicking off "3 spot wiggle" each time and hitting stop as required a violation of principles? It will permit me to do something other than mindlessly staring at the decimal portion of my score. The knowledge gained from evaluation my technique would be about the same with or without the script.

Possible ideas to satisfy the purists:
1. Have different leagues, "manual only" and "script assisted"
2. Add a "total commands issued" to the scoring mechanism to reveal the efficiency/or lack of efficiency, of the solution.

As stated, I am a newbie and some of this may seem a bit naive to more seasoned folders.

Joined: 10/19/2008
Groups: Oma Gawd

Just for your learning on this issue that you note above.

Hotspots don't usually lend themselves to direct reduction via isolated wiggle, due to the high stress in their vicinity, much like a u-joint when two coupled axles are more than N degrees rotated from each other off the common axial line(ASCII - /\). It is often much better to reduce torsion in surrounding areas, even though they appear "green", as this is how the hotspot is relaxed enough to release it's energy by reducing the virtual angle between those two axles... (a visualization construct, not actual physical nature of connection, though imo similar)

Most often, radical manipulation of the area is required to open the hotspot, using rebuild, side chain tweaking, banding, or other more forceful methods, rather than isolated wiggle, which is a mainly meant to be a very low impact manipulation...

FYI...

Joined: 06/17/2010
Topic: Game: Other » Game: Tools
Status: Open » Closed

LUA scripts can do that. Done.

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons