2 replies [Last post]
Joined: 11/14/2009
Groups: None
Time Name Message
10:43 zoran Hi everyone
10:43 mateuszwlodarski hi
10:43 aap2 hello
10:43 Renton Good Morning Brother !
10:43 netolk hi
10:43 Renton new threading looks great btw
10:43 zoran has anyone tried playing with the new selection threading feature released today?
10:44 Renton just started
10:44 aap2 a little bit
10:44 zoran we think it should change the way threading can be explored, but we would like to hear feedback on it soon so that we can refine in time
10:45 Renton It's an amazing feature comepred to the old days, now we can get into the guts of it without having to tiptoe around the basics anymore. 9 out of ten brother !
10:46 zoran good
10:46 aap2 i have some other sugestions for the alligment tool
10:46 aap2 give the templates a name like A,B C etc
10:46 TheGUmmer and please add itto the new interface
10:46 aap2 i currently forget which one i have used
10:46 Renton perhaps a nameable field?
10:47 aap2 also an indication which template im currently using aap2 i currently forget which one i have used
10:46 Renton perhaps a nameable field?
10:47 aap2 also an indication which template im currently using
10:48 aap2 and perhaps an done/not done that shows which templates i have or have not threaded
10:49 aap2 perhaps it should even be a counter
10:51 zoran we have a long list from yesterday, and i think we can address a good chunk of the improvements on it
10:51 Madde_irc I would like a score for parts of the template
10:51 Renton I agree Madde.
10:52 SUCHARD me too
10:52 Renton i kinda thought about the orange and blue line graph above the amino acids in that sequence, to show how many phobic and phillics there are
10:53 Renton so we can move them around and see just how strong each seq was on the outside or inside
10:53 Renton *of the design given, but score would be great.
10:55 TheGUmmer i would like to access the alignments from teh selection interface
10:55 zoran we thought you could get alignments from new interface
10:56 TheGUmmer not that i can find
10:57 Madde_irc no, it's an old interface only feature
10:57 zoran ok, that'll be done quickly
10:57 TheGUmmer woohoo!
10:58 zoran are most people using the new interface now?
10:58 TheGUmmer i am
10:58 mimi difficult to tell I'm not
10:58 SUCHARD most of the time
10:58 aap2 not me, used to the old one
10:59 Dolichwier i still use the old one
10:59 zoran are there other tasks for which old interface is better?
10:59 Madde_irc I'm using the old one only for alignments and local wiggling the ends
10:59 zoran madde: can you not local wiggle in the new?
10:59 Renton i use old one, its still got all the tools in places i can find them quick.
11:00 Madde_irc not exactly that way
11:00 Renton and an unlock of locks segs easily found.
11:00 sonnysu so difficult
11:00 TheGUmmer if select a section in the new and hit W. is it local or global wiggle?
11:00 SUCHARD move little piece of blackbone is easer on old
11:00 TheGUmmer that takes practise in the new
11:01 Madde_irc I think it's global wiggle which affects local
11:01 Renton can the void crushers and Contenders please leave some scraps for the rest of us to rank up ! thanks... hehe. puzzle hogs...
11:02 Seth_Cooper TheGUmmer: it would local wiggle
11:02 TheGUmmer So no selection would be global then
11:03 Madde_irc I think the local wiggle in old is more powerful than in new
11:04 zoran any new suggestions to maximize casp outcomes? especially from those that we haven't heard from yesterday?
11:06 Dolichwier more all hands puzzles
11:06 Renton um, send us heaps of new folders using threads then upload their designs into our group folders
11:07 Renton cannon fodder
11:07 Renton um also have cash prizes. (grin)
11:09 zoran even if we had cash to award, do you think it would help the outcomes?
11:10 Renton i was going to say yes... but realy im just playing the goat.
11:11 zoran i mentioned this yesterday briefly: all hands currently does not seem to diversify the solutions. Most people work on improving the current best solution with little cross-borrowing from other solutions
11:11 Dolichwier ahh, fair enough
11:11 Renton can someone advertise for more folders to come and play so we can increase our teams and solutions to start from
11:11 TheGUmmer could you use them as alignments
11:12 mimi whenever I tried to copy from one soln to another it didn't work
11:12 zoran we will try to change all hands to improve this though
11:12 zoran mimi: how did it fail?
11:12 Christoph I find all hands very difficult, because so may scripts were applied to the high scoring structures that improving them is nearly impossible
11:13 Christoph *many
11:13 mimi just came up with a message saying it didn't work
11:13 TheGUmmer yes the hig scoring puzzles are very tight
11:14 zoran so perhaps doing all hands with earlier structures may be more conducive to mixing
11:14 TheGUmmer yes, or use them as alignments and we can mix and match
11:14 zoran sometimes when we look at the known solutions, it is clear that two or three solutions have gotten different aspects of the protein right
11:14 Christoph copying a part of another solution to an high scoring puzzle will drop the score dramatically and you would have to work for hours to find out if that change would be better
11:14 Renton tight is ok, start searching for new shapes if it gets stuck and has brown bits in it otherwise you have a good solution.
11:15 zoran ... but merging these parts requires lots of restructuring rather than slight refinements
11:15 Renton perhaps making a threaded protein available in the templates in alignments
11:16 Renton (from the high scoring one?
11:16 mimi Its all very "hands-on" work and difficult to fit in with a lot of other ouzzazles happening
11:16 mimi puzzles
11:19 Dolichwier sharing the top alignments/threading instead of the whole protein would be interesting
11:19 zoran ok, we'll try all hands with all candiates in good performing clusters, but neither fully optimized to encourage mixing
11:21 Renton thanks Z and S
11:21 Renton how many solutions do we have for casp 1 and 2 now? any with all green segments yet?
11:22 zoran we're still analyzing the results
11:23 Renton sweet as.
11:24 Renton some high scores this year cmpared to last casp8
11:26 Renton any other exciting news?
11:27 zoran we're busy trying to do all the improvements, and seeing how the results from the first puzzles can guide what we do next
11:28 zoran expect frequent updates this month
11:29 Renton thanks guys, its sometimes hard when we dont have direct contact with you to get stuff out of you to keep us motivated to make the earth move.
11:29 Renton it's been nice to have betahelix coming into the picture more
11:30 zoran yes, next week we'll have David and betahelix on the chat to address some of the protein strategy related questions
11:31 Renton did you mean the Big D ?
11:31 Renton *prof David Baker*
11:31 TheGUmmer no hasselhoff
11:32 zoran yes, "big D" ;)
11:32 Renton YEAH !!!! finally.
11:32 Christoph cool
11:32 CharlieFortsConscience which one brought the Berlin wall down,? I always forget...
11:32 Renton Gummer, KIT is now threading proteins at the university. watch out for turbo boost
11:33 TheGUmmer hehe
11:34 Renton i have to go and buy a pokadot bow tie to wear to the David Baker meeting.
11:35 Renton *tinfoil hat already on standby*
11:35 mib_5gw8gd anything else CASP related?
11:35 Seth_Cooper oops, that was me
11:36 Renton *holds back frothing goat on chain* easy girl... it's just seth.
11:37 Seth_Cooper does the partial threading work alright?
11:37 Renton all jokes aside, we need more time to try out your changes
11:37 Seth_Cooper okay
11:37 Seth_Cooper it hasn't been too long since the update
11:38 Renton we almost need a new beginner puzzle to test it out on or another casp one.
11:38 Madde_irc There are some big casp targets for humans, what is the max. number of residues foldit can deal with properly?
11:38 Seth_Cooper typically we try to limit to < 200
11:39 Seth_Cooper we are planning on trying some of the shorter server targets
11:39 Seth_Cooper so that we can have more!
11:39 aap2 how many casp puzzles are there going to be?
11:39 TheGUmmer per day per week
11:39 aap2 casp proteins
11:39 Christoph I saw there is one with ~65 residues, will we work on that one?
11:40 Renton ide sure like the chance to eat the servers up with the new threading component
11:42 Seth_Cooper most likely we'll post the 65 residue one
11:42 Renton buddaboom
11:43 Renton if we can proove ourselves worthy of our workmanship after this casp results are out, will we go forward with making folding by Human a Reality in the global playground?
11:44 Renton (marketplace)
11:44 CharlieFortsConscience you after a job again aot?
11:44 Renton not me, david hasslehoff.
11:45 Renton and tattoo from fantasy island. The Plane, The Plane...
11:45 Seth_Cooper aap2: we don't really have control over the number of CASP puzzles
11:45 Seth_Cooper it's based on what get released, but you can sort of guess from what's been psoted so far
11:46 Seth_Cooper however, we want to reduce the number of a,b,c by posting more templates in a single puzzle
11:47 Renton who designed the new threading component btw?
11:49 Seth_Cooper we had help from some alignment experts in the Baker Lab (such as tex) and beta_helix and myself worked on it
11:49 Renton fantastic concept !!!
11:49 Seth_Cooper thank you!
11:50 Renton its always refreshing watching people working on the business rather than working in the business. if that makes sense.
11:50 Seth_Cooper I think so
11:50 Seth_Cooper okay, I have a class I need to get to
11:50 Seth_Cooper thanks for chatting again everyone!
11:51 Renton Thanks Seth, Thanks Zoran.

<

Brick's picture
User offline. Last seen 8 weeks 5 days ago. Offline
Joined: 07/15/2008
Groups: Beta Folders
wrong log ?

This looks suspiciously like the log from Wednesday...

Joined: 11/14/2009
Groups: None
fixed it. Thanks, Brick.

fixed it. Thanks, Brick.

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