Does any one know of any examples of computationally designed proteins (i.e. create a sequence to fold to a particular structure) whose predicted structures have been experimentally verified, similar to the paper below?
I am especially looking for examples in the literature of:
1. successful prediction of larger proteins
2. designed sequences with little homology to the known sequence. In other words, I am not looking for the effect of a few mutations on the structure but rather the successful (experimentally verified) prediction of the structure of a protein largely non-homologous to any other protein in the PDB.
3. negative controls - sequences predicted not to fold who have been experimentally verified not to possess tertiary structure
I know of a few examples and am looking for more (this is a good chance to promote your work if you've done this).
http://www.nature.com/nature/journal/v491/n7423/full/nature11600.html
There is a world-wide competition in this field, Critical Assessment of protein Structure Prediction (CASP) .http://predictioncenter.org/.
In CASP, some proteins' structures have been solved but not published, and these proteins' sequences are distributed to many researchers who will build their tertiary structure by computational methods. There are a lot of examples of good and bad predictions. Hope this can help you.
You want to look at the earliest work on protein design, which goes all the way back to the 1980s. Back then a few labs designed simple structures (all helical, or all sheet, for example), made the proteins, and in some cases actually did the structures to verify they were what they were designed to be. The first successes in this area were by Jane Richardson and her group at Duke, and Bill DeGrado (who was in industry back them but is now at Penn). Look up their papers; they are really remarkably visionary. You would also enjoy work by Michael Hecht (Princeton), which takes a slightly different approach to the same problem.
this paper might be helpful for you
http://www.sciencedirect.com/science/article/pii/S0006349512005036
Dear Jeff,
To gauge your expectations I would refer you to the Baker's answer in the discussion that he participated in Science. He explicitly stated that the design and successful confirmation rests on a simple assumption of having sufficiently deep energy minimum, which is equivalent to relatively immobile, rigid structure. The expected success rate is probably not higher than a few percent. This fact simply follows from the contradictory, at first sight, but obvious from the evolutionary point of view, design principles that I cited on other threads. Proteins from the chemical point of view are dead objects, so to acquire attributes of life they need to have contradictory, but necessary components:
1) Structural stability for structural purposes - solid like components,
2) Mobility for functional and adaptation purposes - liquid like components.
Because as we concluded in our paper (Burra PNAS 2009) every protein has a distinct and specific ratio of these components the result is that around 30% of expressable proteins would not fold (or with great difficulty under the right set of constraints), 30% would fold conditionally and only around 30% would fold relatively easily (keep in mind that only around 20% of eukaryotic proteins are self-foldable).
Even if proteins fold easily they still can have hinges. So in reality your question is reduced to : How many foldable units were confirmed experimentally. The answer is "very few", if you look from the point of view of fidelity or "a lot" if you are looking from the point of view of a general fold. A lot, because as the PSI (protein structure initiative, or popularly called structural genomics initiative) showed that folds are becoming scarcer and scarcer which means proteins like what we already know about them.
In essence the answer to your question depends on the level of fidelity and CASP is the best illustration of this principle.
One additional point. Because, in science, we really stretch the truth to get to the "positive" results, a very little discussion can be found about, that there is a bunch of examples of noncolinearity of protein sequence with structure: high sequence homology, low structural similarity. Serpins and prions provide another extreme example of changeable folds (ambivalent sequences). So, in essence, your question has internal limits on answerability.
There are lots of these. A good place to look is david baker's publication list at the university of washington. And brian Kuhlman's page at UNC. The most recent example is the set of ideal protein folds developed in the Baker lab. The classic examples of a truly de novo protein are Kuhlman's TOP7 (which had a novel topology) and Kuhman's janus protein that has two folds.
I have been involved with CASP since 2008 and use fold.it to help design and predict the 3d shapes of proteins... you can find out more at the website *Prof David Baker's Papers are also available through the links pages. http://www.fold.it
Thank you all who answered. I am new to this so I appreciate the help. I found a half dozen or so of the specific type I was looking for (single chain, monomeric proteins greater than 60 residues in length that are almost completely de novo, i.e. not a redesign of the active site of an existing protein). The single chain restriction is partially from the limitations of the software I'm using and partially due to the fact that I believe a functional single chain protein is a more stringent test of the folding algorithim then a functional complex made from simpler subunits.
I am aware of the CASP test. However, I believe there is a subtle difference between the prediction of a solved structure and the prediction of a de novo sequence due to some of the reasons Boguslaw and others have suggested. In the CASP test, it is known that a structure exists. However, the vast majority of protein sequences are unlikely to fold correctly, they are liquid like in Boguslaw's terminology or require negative design in the terminology of the protein design field.
- Badges
- Science topic
- Similar topics
- Papers