Combining machine learning with structure-based protein design to predict and engineer post-translational modifications of proteins.

Post-translational modifications (PTMs) of proteins play a vital role in their function and stability. These modifications influence protein folding, signaling, protein-protein interactions, enzyme activity, binding affinity, aggregation, degradation, and much more. To date, over 400 types of PTMs have been described, representing chemical diversity well beyond the genetically encoded amino acids. Such modifications pose a challenge to the successful design of proteins, but also represent a major opportunity to diversify the protein engineering toolbox. To this end, we first trained artificial neural networks (ANNs) to predict eighteen of the most abundant PTMs, including protein glycosylation, phosphorylation, methylation, and deamidation. In a second step, these models were implemented inside the computational protein modeling suite Rosetta, which allows flexible combination with existing protocols to model the modified sites and understand their impact on protein stability as well as function. Lastly, we developed a new design protocol that either maximizes or minimizes the predicted probability of a particular site being modified. We find that this combination of ANN prediction and structure-based design can enable the modification of existing, as well as the introduction of novel, PTMs. The potential applications of our work include, but are not limited to, glycan masking of epitopes, strengthening protein-protein interactions through phosphorylation, as well as protecting proteins from deamidation liabilities. These applications are especially important for the design of new protein therapeutics where PTMs can drastically change the therapeutic properties of a protein. Our work adds novel tools to Rosetta's protein engineering toolbox that allow for the rational design of PTMs.

Predicting ion mobility collision cross sections using projection approximation with ROSIE-PARCS webserver.

Ion mobility coupled to mass spectrometry informs on the shape and size of protein structures in the form of a collision cross section (CCSIM). Although there are several computational methods for predicting CCSIM based on protein structures, including our previously developed projection approximation using rough circular shapes (PARCS), the process usually requires prior experience with the command-line interface. To overcome this challenge, here we present a web application on the Rosetta Online Server that Includes Everyone (ROSIE) webserver to predict CCSIM from protein structure using projection approximation with PARCS. In this web interface, the user is only required to provide one or more PDB files as input. Results from our case studies suggest that CCSIM predictions (with ROSIE-PARCS) are highly accurate with an average error of 6.12%. Furthermore, the absolute difference between CCSIM and CCSPARCS can help in distinguishing accurate from inaccurate AlphaFold2 protein structure predictions. ROSIE-PARCS is designed with a user-friendly interface, is available publicly and is free to use. The ROSIE-PARCS web interface is supported by all major web browsers and can be accessed via this link (https://rosie.graylab.jhu.edu).

Reliable protein-protein docking with AlphaFold, Rosetta, and replica-exchange.

Despite the recent breakthrough of AlphaFold (AF) in the field of protein sequence-to-structure prediction, modeling protein interfaces and predicting protein complex structures remains challenging, especially when there is a significant conformational change in one or both binding partners. Prior studies have demonstrated that AF-multimer (AFm) can predict accurate protein complexes in only up to 43% of cases.1 In this work, we combine AlphaFold as a structural template generator with a physics-based replica exchange docking algorithm. Using a curated collection of 254 available protein targets with both unbound and bound structures, we first demonstrate that AlphaFold confidence measures (pLDDT) can be repurposed for estimating protein flexibility and docking accuracy for multimers. We incorporate these metrics within our ReplicaDock 2.0 protocol2 to complete a robust in-silico pipeline for accurate protein complex structure prediction. AlphaRED (AlphaFold-initiated Replica Exchange Docking) successfully docks failed AF predictions including 97 failure cases in Docking Benchmark Set 5.5. AlphaRED generates CAPRI acceptable-quality or better predictions for 66% of benchmark targets. Further, on a subset of antigen-antibody targets, which is challenging for AFm (19% success rate), AlphaRED demonstrates a success rate of 51%. This new strategy demonstrates the success possible by integrating deep-learning based architectures trained on evolutionary information with physics-based enhanced sampling. The pipeline is available at github.com/Graylab/AlphaRED.

Stabilizing proteins, simplified: A Rosetta-based webtool for predicting favorable mutations.

Many proteins have low thermodynamic stability, which can lead to low expression yields and limit functionality in research, industrial and clinical settings. This article introduces two, web-based tools that use the high-resolution structure of a protein along with the Rosetta molecular modeling program to predict stabilizing mutations. The protocols were recently applied to three genetically and structurally distinct proteins and successfully predicted mutations that improved thermal stability and/or protein yield. In all three cases, combining the stabilizing mutations raised the protein unfolding temperatures by more than 20°C. The first protocol evaluates point mutations and can generate a site saturation mutagenesis heatmap. The second identifies mutation clusters around user-defined positions. Both applications only require a protein structure and are particularly valuable when a deep multiple sequence alignment is not available. These tools were created to simplify protein engineering and enable research that would otherwise be infeasible due to poor expression and stability of the native molecule.

Ensuring scientific reproducibility in bio-macromolecular modeling via extensive, automated benchmarks.

Each year vast international resources are wasted on irreproducible research. The scientific community has been slow to adopt standard software engineering practices, despite the increases in high-dimensional data, complexities of workflows, and computational environments. Here we show how scientific software applications can be created in a reproducible manner when simple design goals for reproducibility are met. We describe the implementation of a test server framework and 40 scientific benchmarks, covering numerous applications in Rosetta bio-macromolecular modeling. High performance computing cluster integration allows these benchmarks to run continuously and automatically. Detailed protocol captures are useful for developers and users of Rosetta and other macromolecular modeling tools. The framework and design concepts presented here are valuable for developers and users of any type of scientific software and for the scientific community to create reproducible methods. Specific examples highlight the utility of this framework, and the comprehensive documentation illustrates the ease of adding new tests in a matter of hours.

PyRosetta Jupyter Notebooks Teach Biomolecular Structure Prediction and Design.

Biomolecular structure drives function, and computational capabilities have progressed such that the prediction and computational design of biomolecular structures is increasingly feasible. Because computational biophysics attracts students from many different backgrounds and with different levels of resources, teaching the subject can be challenging. One strategy to teach diverse learners is with interactive multimedia material that promotes self-paced, active learning. We have created a hands-on education strategy with a set of sixteen modules that teach topics in biomolecular structure and design, from fundamentals of conformational sampling and energy evaluation to applications like protein docking, antibody design, and RNA structure prediction. Our modules are based on PyRosetta, a Python library that encapsulates all computational modules and methods in the Rosetta software package. The workshop-style modules are implemented as Jupyter Notebooks that can be executed in the Google Colaboratory, allowing learners access with just a web browser. The digital format of Jupyter Notebooks allows us to embed images, molecular visualization movies, and interactive coding exercises. This multimodal approach may better reach students from different disciplines and experience levels as well as attract more researchers from smaller labs and cognate backgrounds to leverage PyRosetta in their science and engineering research. All materials are freely available at https://github.com/RosettaCommons/PyRosetta.notebooks.

Better together: Elements of successful scientific software development in a distributed collaborative community.

Many scientific disciplines rely on computational methods for data analysis, model generation, and prediction. Implementing these methods is often accomplished by researchers with domain expertise but without formal training in software engineering or computer science. This arrangement has led to underappreciation of sustainability and maintainability of scientific software tools developed in academic environments. Some software tools have avoided this fate, including the scientific library Rosetta. We use this software and its community as a case study to show how modern software development can be accomplished successfully, irrespective of subject area. Rosetta is one of the largest software suites for macromolecular modeling, with 3.1 million lines of code and many state-of-the-art applications. Since the mid 1990s, the software has been developed collaboratively by the RosettaCommons, a community of academics from over 60 institutions worldwide with diverse backgrounds including chemistry, biology, physiology, physics, engineering, mathematics, and computer science. Developing this software suite has provided us with more than two decades of experience in how to effectively develop advanced scientific software in a global community with hundreds of contributors. Here we illustrate the functioning of this development community by addressing technical aspects (like version control, testing, and maintenance), community-building strategies, diversity efforts, software dissemination, and user support. We demonstrate how modern computational research can thrive in a distributed collaborative community. The practices described here are independent of subject area and can be readily adopted by other software development communities.

Web-accessible molecular modeling with Rosetta: The Rosetta Online Server that Includes Everyone (ROSIE).

The Rosetta molecular modeling software package provides a large number of experimentally validated tools for modeling and designing proteins, nucleic acids, and other biopolymers, with new protocols being added continually. While freely available to academic users, external usage is limited by the need for expertise in the Unix command line environment. To make Rosetta protocols available to a wider audience, we previously created a web server called Rosetta Online Server that Includes Everyone (ROSIE), which provides a common environment for hosting web-accessible Rosetta protocols. Here we describe a simplification of the ROSIE protocol specification format, one that permits easier implementation of Rosetta protocols. Whereas the previous format required creating multiple separate files in different locations, the new format allows specification of the protocol in a single file. This new, simplified protocol specification has more than doubled the number of Rosetta protocols available under ROSIE. These new applications include pKa determination, lipid accessibility calculation, ribonucleic acid redesign, protein-protein docking, protein-small molecule docking, symmetric docking, antibody docking, cyclic toxin docking, critical binding peptide determination, and mapping small molecule binding sites. ROSIE is freely available to academic users at http://rosie.rosettacommons.org.

Discovery of peptide ligands through docking and virtual screening at nicotinic acetylcholine receptor homology models.

Venom peptide toxins such as conotoxins play a critical role in the characterization of nicotinic acetylcholine receptor (nAChR) structure and function and have potential as nervous system therapeutics as well. However, the lack of solved structures of conotoxins bound to nAChRs and the large size of these peptides are barriers to their computational docking and design. We addressed these challenges in the context of the α4ß2 nAChR, a widespread ligand-gated ion channel in the brain and a target for nicotine addiction therapy, and the 19-residue conotoxin α-GID that antagonizes it. We developed a docking algorithm, ToxDock, which used ensemble-docking and extensive conformational sampling to dock α-GID and its analogs to an α4ß2 nAChR homology model. Experimental testing demonstrated that a virtual screen with ToxDock correctly identified three bioactive α-GID mutants (α-GID[A10V], α-GID[V13I], and α-GID[V13Y]) and one inactive variant (α-GID[A10Q]). Two mutants, α-GID[A10V] and α-GID[V13Y], had substantially reduced potency at the human α7 nAChR relative to α-GID, a desirable feature for α-GID analogs. The general usefulness of the docking algorithm was highlighted by redocking of peptide toxins to two ion channels and a binding protein in which the peptide toxins successfully reverted back to near-native crystallographic poses after being perturbed. Our results demonstrate that ToxDock can overcome two fundamental challenges of docking large toxin peptides to ion channel homology models, as exemplified by the α-GID:α4ß2 nAChR complex, and is extendable to other toxin peptides and ion channels. ToxDock is freely available at rosie.rosettacommons.org/tox_dock.

Computing structure-based lipid accessibility of membrane proteins with mp_lipid_acc in RosettaMP.

BACKGROUND: Membrane proteins are underrepresented in structural databases, which has led to a lack of computational tools and the corresponding inappropriate use of tools designed for soluble proteins. For membrane proteins, lipid accessibility is an essential property. Although programs are available for sequence-based prediction of lipid accessibility and structure-based identification of solvent-accessible surface area, the latter does not distinguish between water accessible and lipid accessible residues in membrane proteins. RESULTS: Here we present mp_lipid_acc, the first method to identify lipid accessible residues from the protein structure, implemented in the RosettaMP framework and available as a webserver. Our method uses protein structures transformed in membrane coordinates, for instance from PDBTM or OPM databases, and a defined membrane thickness to classify lipid accessibility of residues. mp_lipid_acc is applicable to both α-helical and ß-barrel membrane proteins of diverse architectures with or without water-filled pores and uses a concave hull algorithm for surface-residue classification. We further provide a manually curated benchmark dataset that can be used for further method development. CONCLUSIONS: We present a novel tool to classify lipid accessibility from the protein structure, which is applicable to proteins of diverse architectures and achieves prediction accuracies of 90% on a manually curated database. mp_lipid_acc is part of the Rosetta software suite, available at www.rosettacommons.org . The webserver is available at http://rosie.graylab.jhu.edu/mp_lipid_acc/submit and the benchmark dataset is available at http://tinyurl.com/mp-lipid-acc-dataset .

Modeling and docking of antibody structures with Rosetta.

We describe Rosetta-based computational protocols for predicting the 3D structure of an antibody from sequence (RosettaAntibody) and then docking the antibody to protein antigens (SnugDock). Antibody modeling leverages canonical loop conformations to graft large segments from experimentally determined structures, as well as offering (i) energetic calculations to minimize loops, (ii) docking methodology to refine the VL-VH relative orientation and (iii) de novo prediction of the elusive complementarity determining region (CDR) H3 loop. To alleviate model uncertainty, antibody-antigen docking resamples CDR loop conformations and can use multiple models to represent an ensemble of conformations for the antibody, the antigen or both. These protocols can be run fully automated via the ROSIE web server (http://rosie.rosettacommons.org/) or manually on a computer with user control of individual steps. For best results, the protocol requires roughly 1,000 CPU-hours for antibody modeling and 250 CPU-hours for antibody-antigen docking. Tasks can be completed in under a day by using public supercomputers.

Improved prediction of antibody VL-VH orientation.

Antibodies are important immune molecules with high commercial value and therapeutic interest because of their ability to bind diverse antigens. Computational prediction of antibody structure can quickly reveal valuable information about the nature of these antigen-binding interactions, but only if the models are of sufficient quality. To achieve high model quality during complementarity-determining region (CDR) structural prediction, one must account for the VL-VH orientation. We developed a novel four-metric VL-VH orientation coordinate frame. Additionally, we extended the CDR grafting protocol in RosettaAntibody with a new method that diversifies VL-VH orientation by using 10 VL-VH orientation templates rather than a single one. We tested the multiple-template grafting protocol on two datasets of known antibody crystal structures. During the template-grafting phase, the new protocol improved the fraction of accurate VL-VH orientation predictions from only 26% (12/46) to 72% (33/46) of targets. After the full RosettaAntibody protocol, including CDR H3 remodeling and VL-VH re-orientation, the new protocol produced more candidate structures with accurate VL-VH orientation than the standard protocol in 43/46 targets (93%). The improved ability to predict VL-VH orientation will bolster predictions of other parts of the paratope, including the conformation of CDR H3, a grand challenge of antibody homology modeling.

Peptiderive server: derive peptide inhibitors from protein-protein interactions.

The Rosetta Peptiderive protocol identifies, in a given structure of a protein-protein interaction, the linear polypeptide segment suggested to contribute most to binding energy. Interactions that feature a 'hot segment', a linear peptide with significant binding energy compared to that of the complex, may be amenable for inhibition and the peptide sequence and structure derived from the interaction provide a starting point for rational drug design. Here we present a web server for Peptiderive, which is incorporated within the ROSIE web interface for Rosetta protocols. A new feature of the protocol also evaluates whether derived peptides are good candidates for cyclization. Fast computation times and clear visualization allow users to quickly assess the interaction of interest. The Peptiderive server is available for free use at http://rosie.rosettacommons.org/peptiderive.

DARC 2.0: Improved Docking and Virtual Screening at Protein Interaction Sites.

Over the past decade, protein-protein interactions have emerged as attractive but challenging targets for therapeutic intervention using small molecules. Due to the relatively flat surfaces that typify protein interaction sites, modern virtual screening tools developed for optimal performance against "traditional" protein targets perform less well when applied instead at protein interaction sites. Previously, we described a docking method specifically catered to the shallow binding modes characteristic of small-molecule inhibitors of protein interaction sites. This method, called DARC (Docking Approach using Ray Casting), operates by comparing the topography of the protein surface when "viewed" from a vantage point inside the protein against the topography of a bound ligand when "viewed" from the same vantage point. Here, we present five key enhancements to DARC. First, we use multiple vantage points to more accurately determine protein-ligand surface complementarity. Second, we describe a new scheme for rapidly determining optimal weights in the DARC scoring function. Third, we incorporate sampling of ligand conformers "on-the-fly" during docking. Fourth, we move beyond simple shape complementarity and introduce a term in the scoring function to capture electrostatic complementarity. Finally, we adjust the control flow in our GPU implementation of DARC to achieve greater speedup of these calculations. At each step of this study, we evaluate the performance of DARC in a "pose recapitulation" experiment: predicting the binding mode of 25 inhibitors each solved in complex with its distinct target protein (a protein interaction site). Whereas the previous version of DARC docked only one of these inhibitors to within 2 Å RMSD of its position in the crystal structure, the newer version achieves this level of accuracy for 12 of the 25 complexes, corresponding to a statistically significant performance improvement (p < 0.001). Collectively then, we find that the five enhancements described here--which together make up DARC 2.0--lead to dramatically improved speed and performance relative to the original DARC method.

Adding diverse noncanonical backbones to rosetta: enabling peptidomimetic design.

Peptidomimetics are classes of molecules that mimic structural and functional attributes of polypeptides. Peptidomimetic oligomers can frequently be synthesized using efficient solid phase synthesis procedures similar to peptide synthesis. Conformationally ordered peptidomimetic oligomers are finding broad applications for molecular recognition and for inhibiting protein-protein interactions. One critical limitation is the limited set of design tools for identifying oligomer sequences that can adopt desired conformations. Here, we present expansions to the ROSETTA platform that enable structure prediction and design of five non-peptidic oligomer scaffolds (noncanonical backbones), oligooxopiperazines, oligo-peptoids, [Formula: see text]-peptides, hydrogen bond surrogate helices and oligosaccharides. This work is complementary to prior additions to model noncanonical protein side chains in ROSETTA. The main purpose of our manuscript is to give a detailed description to current and future developers of how each of these noncanonical backbones was implemented. Furthermore, we provide a general outline for implementation of new backbone types not discussed here. To illustrate the utility of this approach, we describe the first tests of the ROSETTA molecular mechanics energy function in the context of oligooxopiperazines, using quantum mechanical calculations as comparison points, scanning through backbone and side chain torsion angles for a model peptidomimetic. Finally, as an example of a novel design application, we describe the automated design of an oligooxopiperazine that inhibits the p53-MDM2 protein-protein interaction. For the general biological and bioengineering community, several noncanonical backbones have been incorporated into web applications that allow users to freely and rapidly test the presented protocols (http://rosie.rosettacommons.org). This work helps address the peptidomimetic community's need for an automated and expandable modeling tool for noncanonical backbones.

Alternative computational protocols for supercharging protein surfaces for reversible unfolding and retention of stability.

Reengineering protein surfaces to exhibit high net charge, referred to as "supercharging", can improve reversibility of unfolding by preventing aggregation of partially unfolded states. Incorporation of charged side chains should be optimized while considering structural and energetic consequences, as numerous mutations and accumulation of like-charges can also destabilize the native state. A previously demonstrated approach deterministically mutates flexible polar residues (amino acids DERKNQ) with the fewest average neighboring atoms per side chain atom (AvNAPSA). Our approach uses Rosetta-based energy calculations to choose the surface mutations. Both protocols are available for use through the ROSIE web server. The automated Rosetta and AvNAPSA approaches for supercharging choose dissimilar mutations, raising an interesting division in surface charging strategy. Rosetta-supercharged variants of GFP (RscG) ranging from -11 to -61 and +7 to +58 were experimentally tested, and for comparison, we re-tested the previously developed AvNAPSA-supercharged variants of GFP (AscG) with +36 and -30 net charge. Mid-charge variants demonstrated ∼3-fold improvement in refolding with retention of stability. However, as we pushed to higher net charges, expression and soluble yield decreased, indicating that net charge or mutational load may be limiting factors. Interestingly, the two different approaches resulted in GFP variants with similar refolding properties. Our results show that there are multiple sets of residues that can be mutated to successfully supercharge a protein, and combining alternative supercharge protocols with experimental testing can be an effective approach for charge-based improvement to refolding.

Serverification of molecular modeling applications: the Rosetta Online Server that Includes Everyone (ROSIE).

The Rosetta molecular modeling software package provides experimentally tested and rapidly evolving tools for the 3D structure prediction and high-resolution design of proteins, nucleic acids, and a growing number of non-natural polymers. Despite its free availability to academic users and improving documentation, use of Rosetta has largely remained confined to developers and their immediate collaborators due to the code's difficulty of use, the requirement for large computational resources, and the unavailability of servers for most of the Rosetta applications. Here, we present a unified web framework for Rosetta applications called ROSIE (Rosetta Online Server that Includes Everyone). ROSIE provides (a) a common user interface for Rosetta protocols, (b) a stable application programming interface for developers to add additional protocols, (c) a flexible back-end to allow leveraging of computer cluster resources shared by RosettaCommons member institutions, and (d) centralized administration by the RosettaCommons to ensure continuous maintenance. This paper describes the ROSIE server infrastructure, a step-by-step 'serverification' protocol for use by Rosetta developers, and the deployment of the first nine ROSIE applications by six separate developer teams: Docking, RNA de novo, ERRASER, Antibody, Sequence Tolerance, Supercharge, Beta peptide design, NCBB design, and VIP redesign. As illustrated by the number and diversity of these applications, ROSIE offers a general and speedy paradigm for serverification of Rosetta applications that incurs negligible cost to developers and lowers barriers to Rosetta use for the broader biological community. ROSIE is available at http://rosie.rosettacommons.org.

Scientific benchmarks for guiding macromolecular energy function improvement.

Accurate energy functions are critical to macromolecular modeling and design. We describe new tools for identifying inaccuracies in energy functions and guiding their improvement, and illustrate the application of these tools to the improvement of the Rosetta energy function. The feature analysis tool identifies discrepancies between structures deposited in the PDB and low-energy structures generated by Rosetta; these likely arise from inaccuracies in the energy function. The optE tool optimizes the weights on the different components of the energy function by maximizing the recapitulation of a wide range of experimental observations. We use the tools to examine three proposed modifications to the Rosetta energy function: improving the unfolded state energy model (reference energies), using bicubic spline interpolation to generate knowledge-based torisonal potentials, and incorporating the recently developed Dunbrack 2010 rotamer library (Shapovalov & Dunbrack, 2011).

Real-time PyMOL visualization for Rosetta and PyRosetta.

Computational structure prediction and design of proteins and protein-protein complexes have long been inaccessible to those not directly involved in the field. A key missing component has been the ability to visualize the progress of calculations to better understand them. Rosetta is one simulation suite that would benefit from a robust real-time visualization solution. Several tools exist for the sole purpose of visualizing biomolecules; one of the most popular tools, PyMOL (Schrödinger), is a powerful, highly extensible, user friendly, and attractive package. Integrating Rosetta and PyMOL directly has many technical and logistical obstacles inhibiting usage. To circumvent these issues, we developed a novel solution based on transmitting biomolecular structure and energy information via UDP sockets. Rosetta and PyMOL run as separate processes, thereby avoiding many technical obstacles while visualizing information on-demand in real-time. When Rosetta detects changes in the structure of a protein, new coordinates are sent over a UDP network socket to a PyMOL instance running a UDP socket listener. PyMOL then interprets and displays the molecule. This implementation also allows remote execution of Rosetta. When combined with PyRosetta, this visualization solution provides an interactive environment for protein structure prediction and design.

ROSETTA3: an object-oriented software suite for the simulation and design of macromolecules.

We have recently completed a full re-architecturing of the ROSETTA molecular modeling program, generalizing and expanding its existing functionality. The new architecture enables the rapid prototyping of novel protocols by providing easy-to-use interfaces to powerful tools for molecular modeling. The source code of this rearchitecturing has been released as ROSETTA3 and is freely available for academic use. At the time of its release, it contained 470,000 lines of code. Counting currently unpublished protocols at the time of this writing, the source includes 1,285,000 lines. Its rapid growth is a testament to its ease of use. This chapter describes the requirements for our new architecture, justifies the design decisions, sketches out central classes, and highlights a few of the common tasks that the new software can perform.