Cover the Ligand
As environmental impact becomes an increasing concern for chemical manufacturing, the idea of using enzymes in environmentally friendly conditions to catalyze chemical transformations instead of traditional chemistry is becoming increasingly important. Unfortunately, naturally occurring enzymes only catalyze a subset of the reactions that are readily available to the synthetic chemist. Therefore engineering enzymes to catalyze novel chemical reactions is an area of great interest. To address this we are using the computational tool ROSETTA to engineer an enzyme capable of catalyzing one such reaction: an intermolecular, stereospecific Diels-Alder reaction. While no clear examples of this reaction have been shown to occur in nature, it is one of the corner stone reactions used by synthetic chemists to form carbon-carbon bonds and received a Nobel prize in 1950 (http://en.wikipedia.org/wiki/Diels%E2%80%93Alder_reaction). While popular with synthetic chemists, this reaction often falters due to the large number of different products that can be formed when bringing two molecules together. With an enzyme we can control the way the two molecules come together during the reaction, therefore we will be able to catalyze the formation of a single product. We hope that this novel enzyme and more like it will introduce a new class of environmentally friendly chemical reactions available to the biochemist’s tool bench and lead to the creation of new drugs, commodity chemicals, and more!
The “Cover the Ligand” Puzzle is one of the Diels-Alder enzymes we are currently engineering. Unfortunately only half of the ligand is covered making it so our initial design is a weak binder. With your help we would like to increase the size of the loops surrounding the ligand, allowing additional amino acids to touch the ligand, which should result in an increase of this enzymes catalytic efficiency.
Your designs will be ranked based on score and number of additional contacts made to the ligand. The most promising of your designs will then be synthesized and experimentally tested in the Baker lab. We look forward to seeing all of the exciting new solutions you come up with!( Posted by siegeljb 73 1944 | Tue, 08/25/2009 - 21:15 | 2 comments )
Foldit scripting preview
We've started working on adding Lua scripting, and here's a very rough preview of what it looks like now. This is just in a command line window, but there will also be support for scripts in the cookbook.
What do you think? What kind of API would you like to see? Let us know in the comments!( Posted by Seth Cooper 73 1944 | Fri, 08/21/2009 - 21:04 | 10 comments )
We recently added an option to turn the outlines on the protein structure on and off. This was initially for performance reasons, but it has also come up that the protein may look better without them as well. So, here's a poll to decide what we should do with the outlines, let us know what you think!( Posted by Seth Cooper 73 1944 | Fri, 08/21/2009 - 19:07 | 1 comment )
Design of inhibitors update
The results from the inhibitor design puzzle have come in and it appears that many of the players fully identified all of the hot-spot residues - those are the residues on the inhibitor that are crucial to inhibition in the natural inhibitor. This is very impressive because getting these hot-spot residues correctly requires a combination of tools that Rosetta provides, including sidechain redesign and minimization.
We will next provide several other puzzles similar to this one, where we know what the hot-spot residue identities are, but we shaved those off and you can try to recapitulate them. The first in this set of puzzles is the structure of human growth hormone bound to its receptor.
Following a few of these puzzles we would like to ask for your help in improving our designs for a variety of protein targets. I'm looking forward to your participation and am eager to see what design possibilities you could uncover that we miss.( Posted by sarelf 73 1944 | Tue, 08/04/2009 - 23:30 | 0 comments )
Mini CASP-like competition starts today
We have exciting news!
We were able to obtain sequences for proteins that are currently
unsolved, but will be released to the Protein Data Bank very soon!
In fact, today's first puzzle will be released to the PDB next week,
so we need you to start folding quickly!
After the first puzzle (released today) expires on August 8th, we will
release another puzzle where you will have more time. The second
protein will not be deposited into the PDB until the end of the month.
We will give you starting models outputted from Rosetta@home, and
since we have no idea which predictions are correct we will try to
give you as diverse a set as possible.
We are very excited to be able to run this blind experiment where the
proteins will be released so soon!
This will be perfect practice for everyone as CASP9 starts next May,
and it will help us figure out what are the best Rosetta models to use
as starting points for Foldit.