Hi everybody! A quick announcement today- our next Science Chat is March 28. Be sure your questions get posted here (this thread, right here) so I can send them to the team in advance of the chat. As long time chat fans are aware, the more questions we have prepared ahead of time the more questions we can likely get in. Keep in mind there were a lot of questions for our last chat (which we appreciated so much) and if we don't get to your question during the chat (it is unlikely we will be able to address development related questions during this hour, although you're free to ask them and we will endeavor to get an answer later), it's entirely possible you may see it in a future blog.
The Date: Tuesday, 28 March 2017
The location: #veteran, IRC
The Time: 2100-2200 (or so, but the official chat ends after an hour) GMT (aka 1400-1500 PT)
The Time Zone Converter: Right this way!
We look forward to discussing science with you!
http://fold.it/portal/node/2003481 should have what you're looking for :)
Ignoring the filter score in design puzzles what are the differences in the scoring function between design puzzles and prediction puzzles? It seems that the backbone score is significantly lower for segments lying outside their rama areas in design puzzles versus prediction puzzles. Also I've noticed that the top score for "unsolved denovo freestyle puzzles" (1) based on a foldit design in generally higher that the same score for the original foldit design (2) (substracting the filter bonus), and that the dots in the rama map for (1) is more spread out than they were in (2).
Do you ever take an interesting backbone from a foldit design and use Rosetta to find better sidechains for it?
Why can't the Blueprint tool be enabled on all puzzles?
One of the big themes to a recent share from the developers is that they really want to help us design good loops (close to ideality). Applying the BP tool to a loop helps to constrain its shape while you do other things to the tertiary structure.
The Rama Map is hard to use because it represents everything in stick mode. I can't tell where the SS's end and the loop begins. But it does offer some interesting ways to build your tertiary structure.
Shawn
Why aren't you giving the prediction scores for the SS any longer? The recent sequence of 1344, 1347, 1350 didn't list the confidences for the SS predictions.
Having those prediction confidences, along with the Alignment tool (for 1347 and 1350) may have helped us get closer to a real solution to this protein. For 1347 and 1350, selecting slightly randomized (detuned) versions of the top 5 best scoring general shapes would have benefited everyone. This is how the you in the lab can collaborate with us (the explorers) to drive towards a native solution more quickly. You could call this the "scientist filter". That way, if you don't have a team to evo from or you didn't play the earlier puzzle, you still get some starting points.
Have you made any progress with improving the stability of the client in symmetry puzzles?
