[NewChapter] Frozen on terminal end

Case number:954892-996665
Topic:Developer Preview
Opened by:auntdeen
Opened on:Thursday, January 16, 2014 - 05:31
Last modified:Tuesday, December 29, 2020 - 11:58

I've been playing around with the Spider Toxin and noticed that doing rebuilds anywhere on the protein, when wiggling out, the last segment sticks in place while the rest of the protein moves. This happens even when rebuilding that end - the rebuild seems "frozen" instead of having any movement during wiggle.

That end will rebuild, even into a helix - but even when modified, wiggle affects the rest of the protein but not the last one or two segments. This is all visual. Comparing in the Main client, that end is not "frozen" in space while wiggling.

If you freeze off the last segment and the first segment and do a local wiggle, the first segment will increase in score at least some fractions. The last segment will not budge even a hundredth, even on a fresh reset of the puzzle.

(Thu, 01/16/2014 - 05:31  |  7 comments)

Joined: 04/19/2009

Today I've played around with the Mini newchapter puzzle and confirmed what I saw yesterday. I simply made most of it into a helix with loop ends 4 segments on either end. If the terminal end isn't, by luck, the first to curl up into shape, the lock on the last seg - or segs, I am still not sure how many get involved with rebuild - makes it almost impossible to get a good helix shape overall.

This bug may have major implications, and is likely contributing to the reason why rebuild scripts are performing so poorly in the newchapter client.

Joined: 04/19/2009

I've been running an Ebola Rebuild for a few hours.

The script keeps focusing on that terminal end (as it should) trying to rebuild it. It simply isn't working, and the results suggest that any drw script is possibly getting hung up in an area that will not accept rebuild because of the locked terminal segment.

wisky's picture
User offline. Last seen 1 year 11 weeks ago. Offline
Joined: 07/13/2011

I ran through every newchapter puzzle, and put a band between the last segment and a fairly far away segment, then wiggled it. In every case, the last segment stays perfectly locked in place (It doesn't even appear to rotate), and the next 4-5 segments down the line move very little. With bands in space, the whole protein moves to the space where the band is, but the terminal end stays locked.

Only using local wiggle can I get the terminal end to move.

NickyCGS's picture
User offline. Last seen 6 years 37 weeks ago. Offline
Joined: 06/27/2013
Groups: Repro-men

Thanks for the feedback Auntdeen and Wisky. Reproduced this issue and sent to developers.

Joined: 04/19/2009

Thanks Nicky

For more info on when we have encountered this issue in the past (but it was middle segments locking) please see this feedback, which contains links to others with the same bug:


And this one from an earlier time:


Although it was never completely clear whether this was a nasty bug or a feature, because this affects the terminal end so badly (and scripts), in this case it is a very nasty bug :(

Joined: 11/10/2007
Groups: Window Group

This should be fixed in the next update to newchapter.

Joined: 09/24/2012
Groups: Go Science
Status: Open » Closed



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, Boehringer Ingelheim, RosettaCommons