Macromolecular Modeling Blog ™

In a recent review published in Biochemistry, Kaufmann, Lemmon and DeLuca et al. from the Meiler Lab, cover the Rosetta modeling suite capabilities. More importantly, as supplementary information they provide tutorials to demonstrate 6 basic use cases of Rosetta.

(more…)

This is the fifth and last post in the CAPRI series, summarizing the presentations of Xiaoqin Zou and Ora Schueler-Furman (Saving the best for last..), as provided by the speakers. I hope the CAPRI series was able to give a snapshot of the state of computational protein-protein docking and its community. I want to thank again to everyone that took part in the meeting and helped me with this series.

(more…)

As in previous years, we would like to invite representatives of the pharma and biotech industry to participate in RosettaCon 2010. Take part in the workshops, seminars, and have an opportunity to meet the Rosetta Commons researchers. Get to know Rosetta from the inside. RosettaCon will take place during August 3-6 at the beautiful Sleeping Lady Mountain Retreat in Leavenworth, Washington, about 2.5 hours outside Seattle, in a dramatic mountain scenery – the only fitting environment for a scientific discussion really.

(more…)

PyRosetta Book

The Gray Lab at Johns Hopkins University has just released PyRosetta, a Python-based interactive platform for accessing the objects and algorithms within the Rosetta protein structure prediction suite.

In addition to the code, the Gray Lab has put together a book that leads the reader through basics of protein structure and energetics to applications in folding, refinement, docking and design. The focus is on enabling users to write custom scripts, so it includes material on Rosetta fundamentals and the appendices have a list of PyRosetta commands and a breakdown of the input files. The book was beta-tested by students during a course at JHU. The course is a series of workshops that teach how to measure and manipulate protein conformations, calculate energies in low- and high-resolution representations, fold proteins from sequence, model variable regions of proteins (loops), dock proteins or small molecules, design protein sequences, and build custom protocols for operations tailored to particular biomolecular applications

The book can be purchased through Lulu:

http://www.lulu.com/content/paperback-book/the-pyrosetta-interactive-platform-for-protein-structure-prediciton-and-design-a-set-of-educational-modules/7187010

or downloaded for free as pdf chapters from http://www.pyrosetta.org under the Tutorial link.

RosettaCon 2009 has ended just a few days ago.

This was a very exciting meeting, gathering almost all Rosetta developers from around the world, and representatives of the major industry licensees of Rosetta. The talks were quite diverse, ranging from new Rosetta protocols under development, applications with regards to numerous biological systems, and up to code architecture and future development of the code.

For scientists specializing in a specific field (such as computational structural biology – just off the top of my head) there are rare occasions in which meetings are dedicated entirely to their topic. This was such a conference! All through the three days of conference and one day of hiking, people had talked, discussed, brain stormed, ate, breathed, thought, swam, hiked, mingled and laughed protein structure. If the coffee break was drawn as a comic figure, with bubble clouds to depict thoughts and conversations, it would be a sea of alpha-helices, weird looking loops, rotamers, ligands and co-factors at interfaces, strange folds and stranger rainbow colored cartoon proteins. A modeler’s utopia.
(more…)

In a post yesterday on the Bleeding Edge Biotech Blog, Adam Kraut gives an expanded version of his article for Bio-IT World entitled Antibody Docking on the Amazon Cloud describing how Pfizer is leveraging the power and flexibility of cloud computing to run antibody docking simulations using Rosetta. Pfizer employed BioTeam (which Kraut consults for) to port Rosetta to the cloud. This is a fascinating article and certainly the Pfizer/BioTeam/Rosetta/Amazon synergy is a phenomenon on the bleeding edge.

I would add to the article that Jeffrey Gray at Johns Hopkins University was the lead developer of RosettaDock and RosettaAntibody. Rosetta has become synonymous with David Baker (and for good reason), but it is important to acknowledge continuing contributions to Rosetta outside of the Baker lab. Scientific collaborations can dissolve quickly when recognition is unevenly distributed, whether internally or as a result of outside perceptions (i.e. not as the result of any intentions by researchers in the collaboration).

If you use or write about Rosetta, please help us keep this unique collaboration going by acknowledging and citing the individual developers and researchers and their current labs!!!!

Next week, RDG will be teaching a three day course at the Center of Information Technology of the National Institutes of Health:

#468 Rosetta Hands-On Workshop

We have recently conducted a poll amongst people interested in learning more about the Rosetta software (as a preliminary step to the Rosetta Academic Training Workshop). One of the questions in that poll was: “Which Rosetta related topics is of the most interest to you?” The results (from ~200 participants) are summarized in the graph below.

Interest trends amongst Rosetta users

Protein-protein docking was chosen as the topic with the highest interest level, with a slight gap from ab-initio structure prediction and protein-ligand docking. This correlates well with another superficial analysis we made a while back (What’s trendy in structural bioinformatics) in which we showed qualitatively that docking receives the most interest from ‘Cite U Like‘ structural bioinformaticians. It is worth noting that Molecular Dynamics was not included, since Rosetta is not intended for MD simulations.

What is your field in structural bioinformatics? Anyone has more evidence for the increase in interest in protein-protein docking?

Rosetta 3.0 was release on the Rosetta Commons website last Friday, February 27th.  Rosetta 3.0 is a whole new Rosetta developed in a purely object oriented manner.  Most functionality from previous versions of Rosetta have been ported over and tested, ensuring very similar results.  Rosetta 3.0 is available for download as a BETA version and any users are encouraged to give feedback.

A big improvement over previous versions of Rosetta is the ability to create custom “apps”, as well as compile existing and documented apps as separate executables.  Gone are the days of a single executable all modes with a paragraph worth of flags.  Now, each app has its own executable and flags.

Existing apps that can be compiled and run with no modifications include:

RosettaAbinitio – Performs de novo protein structure prediction.

RosettaDesign – Dentifies low free energy sequences for target protein backbones.

RosettaDesign pymol plugin – A user-friendly interface for submitting Protein Design simulations using RosettaDesign.

RosettaDock – Predicts the structure of a protein-protein complex from the individual structures of the monomer components.

RosettaAntibody – Predicts antibody Fv region structures and performs antibody-antigen docking.

RosettaFragments – Generates fragment libraries for use by Rosetta ab initio in building protein structures.

RosettaNMR – Incorporates NMR data into the basic Rosetta protocol to accelerate the process of NMR structure prediction.

RosettaDNA – For the design of proteins that interact with specified DNA sequences.

RosettaRNA – Fragment assembly of RNA.

RosettaLigand – For small molecule – protein docking.

It has become popular to say that ab-initio prediction of protein structure is now unnecessary since the “Fold space” is nearly covered and thus there is a representative, homologous solved structure, for every protein. Therefore, limiting the structure prediction problem to homology modeling. However, this is certainly not the case for membrane proteins.

Experimental determination of high-resolution membrane protein structures remains very difficult. The fact that membrane proteins are typically longer than 200 aa does not make the problem easier. Membrane proteins can be classified into 2 groups: transmembrane helical (TMH) bundles and beta-barrels. For TMH proteins, the physical constraints imposed by the anisotropic environment of the lipid bilayer lead to characteristic distributions of amino acids that depend on their depth in the membrane. These observations have enabled the development of topology prediction schemes that have become quite sophisticated and powerful. 

The authors expend the technique developed for sampling nonlocal beta-sheet topologies to fold membrane proteins from sequence. In this scheme, the relative orientation of TMH pairs is fixed at two particular positions during folding by long-range pair wise constraints. For each constraint between two helices, a ‘‘fold tree’’ is constructed for the polypeptide chain in which two C-alpha positions from the two helices are connected and fixed in space during folding. To allow for this non-local connection in the tree, the peptide chain is cut elsewhere between the two connected positions. The cut is randomly selected within predicted loop regions of the proteins with a bias toward long loops. This avoids disrupting subdomains composed of few TMHs connected by short loops, which can be folded properly. 

To predict those structural constraints from sequence information, a database of  TMHs configurations from TMH pairs of known structures was assembled. This database of interacting TMH pairs is searched for local sequence matches with all possible pairs of predicted TMHs in the query sequence using a sliding window. In each folding trajectory, a single randomly selected predicted interaction in the library is used to constrain a particular helix pair to the helix–helix arrangement. Ten predicted interactions are included for each helix pair, which allows correct models to be generated despite the low overall accuracy of the interaction library since only one of the 10 is requiered to be correct.

The method was tested on 12 membrane proteins of diverse topologies and functions with lengths ranging between 190 and 300 residues. Enforcing a single constraint during the folding simulations enriched the population of near-native models for 9 proteins over the predictions made with the older generation of RosettaMembrane. In 4 of the cases in which the constraint was predicted from the sequence, 1 of the 5 lowest energy models was superimposable within 4 Å on the native structure. Near-native structures could also be selected for heme-binding and pore-forming domains from simulations in which pairs of conserved histidine-chelating hemes and one experimentally determined salt bridge were constrained, respectively. In 8 out of the 12 cases a model was sampled in which more than 85% of the sequence was superimposable onto the native structure and in 5 cases this was true for one of the 5 lowest energy models.

P. Barth, B. Wallner, D. Baker (2009). Prediction of membrane protein structures with complex topologies using limited constraints Proceedings of the National Academy of Sciences, 106 (5), 1409-1414 DOI: 10.1073/pnas.0808323106