Improving the accuracy of the structure prediction of the third hypervariable loop of the heavy chains of antibodies

Computational Modeling of Antibody and T-Cell Receptor (CDR3 Loops)

Antibodies and T-cell receptors have been a subject of much interest due to their central role in the immune system and their potential applications in several biotechnological and medical applications from cancer therapy to vaccine development. A unique feature of these two lymphocyte receptors is their ability to bind a huge variety of different (pathogen) targets. This ability stems from six short loops in the binding domain that have hypervariable sequence due to genetic recombination mechanism. Particularly one of these loops, the third complementarity determining region (CDR3), has the highest sequence variability and a dominant role in binding the target. However, it has also been proven the most difficult to be modeled structurally, which is vitally important for downstream tasks such as binding prediction. This difficulty stems from its variability in sequence that both reduces the possibility of finding homologues and introduces unique structural features in the loop. We present here a general protocol for modeling such loops in antibodies and T-cell receptors. We also discuss the difficulties in loop modeling and the advantages and limitations of different modeling methods.

Quantitative Description of Surface Complementarity of Antibody-Antigen Interfaces

Antibodies have the remarkable ability to recognise their cognate antigens with extraordinary affinity and specificity. Discerning the rules that define antibody-antigen recognition is a fundamental step in the rational design and engineering of functional antibodies with desired properties. In this study we apply the 3D Zernike formalism to the analysis of the surface properties of the antibody complementary determining regions (CDRs). Our results show that shape and electrostatic 3DZD descriptors of the surface of the CDRs are predictive of antigen specificity, with classification accuracy of 81% and area under the receiver operating characteristic curve (AUC) of 0.85. Additionally, while in terms of surface size, solvent accessibility and amino acid composition, antibody epitopes are typically not distinguishable from non-epitope, solvent-exposed regions of the antigen, the 3DZD descriptors detect significantly higher surface complementarity to the paratope, and are able to predict correct paratope-epitope interaction with an AUC = 0.75.

Quality Assessment of Protein Docking Models Based on Graph Neural Network

Protein docking provides a structural basis for the design of drugs and vaccines. Among the processes of protein docking, quality assessment (QA) is utilized to pick near-native models from numerous protein docking candidate conformations, and it directly determines the final docking results. Although extensive efforts have been made to improve QA accuracy, it is still the bottleneck of current protein docking systems. In this paper, we presented a Deep Graph Attention Neural Network (DGANN) to evaluate and rank protein docking candidate models. DGANN learns inter-residue physio-chemical properties and structural fitness across the two protein monomers in a docking model and generates their probabilities of near-native models. On the ZDOCK decoy benchmark, our DGANN outperformed the ranking provided by ZDOCK in terms of ranking good models into the top selections. Furthermore, we conducted comparative experiments on an independent testing dataset, and the results also demonstrated the superiority and generalization of our proposed method.

Computational approaches to therapeutic antibody design: established methods and emerging trends

Antibodies are proteins that recognize the molecular surfaces of potentially noxious molecules to mount an adaptive immune response or, in the case of autoimmune diseases, molecules that are part of healthy cells and tissues. Due to their binding versatility, antibodies are currently the largest class of biotherapeutics, with five monoclonal antibodies ranked in the top 10 blockbuster drugs. Computational advances in protein modelling and design can have a tangible impact on antibody-based therapeutic development. Antibody-specific computational protocols currently benefit from an increasing volume of data provided by next generation sequencing and application to related drug modalities based on traditional antibodies, such as nanobodies. Here we present a structured overview of available databases, methods and emerging trends in computational antibody analysis and contextualize them towards the engineering of candidate antibody therapeutics.

Characterizing Epitope Binding Regions of Entire Antibody Panels by Combining Experimental and Computational Analysis of Antibody: Antigen Binding Competition

Vaccines and immunotherapies depend on the ability of antibodies to sensitively and specifically recognize particular antigens and specific epitopes on those antigens. As such, detailed characterization of antibody-antigen binding provides important information to guide development. Due to the time and expense required, high-resolution structural characterization techniques are typically used sparingly and late in a development process. Here, we show that antibody-antigen binding can be characterized early in a process for whole panels of antibodies by combining experimental and computational analyses of competition between monoclonal antibodies for binding to an antigen. Experimental "epitope binning" of monoclonal antibodies uses high-throughput surface plasmon resonance to reveal which antibodies compete, while a new complementary computational analysis that we call "dock binning" evaluates antibody-antigen docking models to identify why and where they might compete, in terms of possible binding sites on the antigen. Experimental and computational characterization of the identified antigenic hotspots then enables the refinement of the competitors and their associated epitope binding regions on the antigen. While not performed at atomic resolution, this approach allows for the group-level identification of functionally related monoclonal antibodies (i.e., communities) and identification of their general binding regions on the antigen. By leveraging extensive epitope characterization data that can be readily generated both experimentally and computationally, researchers can gain broad insights into the basis for antibody-antigen recognition in wide-ranging vaccine and immunotherapy discovery and development programs.

How repertoire data are changing antibody science

Antibodies are vital proteins of the immune system that recognize potentially harmful molecules and initiate their removal. Mammals can efficiently create vast numbers of antibodies with different sequences capable of binding to any antigen with high affinity and specificity. Because they can be developed to bind to many disease agents, antibodies can be used as therapeutics. In an organism, after antigen exposure, antibodies specific to that antigen are enriched through clonal selection, expansion, and somatic hypermutation. The antibodies present in an organism therefore report on its immune status, describe its innate ability to deal with harmful substances, and reveal how it has previously responded. Next-generation sequencing technologies are being increasingly used to query the antibody, or B-cell receptor (BCR), sequence repertoire, and the amount of BCR data in public repositories is growing. The Observed Antibody Space database, for example, currently contains over a billion sequences from 68 different studies. Repertoires are available that represent both the naive state (i.e. antigen-inexperienced) and that after immunization. This wealth of data has created opportunities to learn more about our immune system. In this review, we discuss the many ways in which BCR repertoire data have been or could be exploited. We highlight its utility for providing insights into how the naive immune repertoire is generated and how it responds to antigens. We also consider how structural information can be used to enhance these data and may lead to more accurate depictions of the sequence space and to applications in the discovery of new therapeutics.

Modeling Antibody-Antigen Complexes by Information-Driven Docking

Antibodies are Y-shaped proteins essential for immune response. Their capability to recognize antigens with high specificity makes them excellent therapeutic targets. Understanding the structural basis of antibody-antigen interactions is therefore crucial for improving our ability to design efficient biological drugs. Computational approaches such as molecular docking are providing a valuable and fast alternative to experimental structural characterization for these complexes. We investigate here how information about complementarity-determining regions and binding epitopes can be used to drive the modeling process, and present a comparative study of four different docking software suites (ClusPro, LightDock, ZDOCK, and HADDOCK) providing specific options for antibody-antigen modeling. Their performance on a dataset of 16 complexes is reported. HADDOCK, which includes information to drive the docking, is shown to perform best in terms of both success rate and quality of the generated models in both the presence and absence of information about the epitope on the antigen.

SWISS-MODEL: homology modelling of protein structures and complexes

Homology modelling has matured into an important technique in structural biology, significantly contributing to narrowing the gap between known protein sequences and experimentally determined structures. Fully automated workflows and servers simplify and streamline the homology modelling process, also allowing users without a specific computational expertise to generate reliable protein models and have easy access to modelling results, their visualization and interpretation. Here, we present an update to the SWISS-MODEL server, which pioneered the field of automated modelling 25 years ago and been continuously further developed. Recently, its functionality has been extended to the modelling of homo- and heteromeric complexes. Starting from the amino acid sequences of the interacting proteins, both the stoichiometry and the overall structure of the complex are inferred by homology modelling. Other major improvements include the implementation of a new modelling engine, ProMod3 and the introduction a new local model quality estimation method, QMEANDisCo. SWISS-MODEL is freely available at https://swissmodel.expasy.org.

PIGSPro: prediction of immunoGlobulin structures v2

PIGSpro is a significant upgrade of the popular PIGS server for the prediction of the structure of immunoglobulins. The software has been completely rewritten in python following a similar pipeline as in the original method, but including, at various steps, relevant modifications found to improve its prediction accuracy, as demonstrated here. The steps of the pipeline include the selection of the appropriate framework for predicting the conserved regions of the molecule by homology; the target template alignment for this portion of the molecule; the selection of the main chain conformation of the hypervariable loops according to the canonical structure model, the prediction of the third loop of the heavy chain (H3) for which complete canonical structures are not available and the packing of the light and heavy chain if derived from different templates. Each of these steps has been improved including updated methods developed along the years. Last but not least, the user interface has been completely redesigned and an automatic monthly update of the underlying database has been implemented. The method is available as a web server at http://biocomputing.it/pigspro.

High-Throughput Antibody Structure Modeling and Design Using ABodyBuilder

Antibodies are proteins of the adaptive immune system; they can be designed to bind almost any molecule, and are increasingly being used as biotherapeutics. Experimental antibody design is an expensive and time-consuming process, and computational antibody design methods can now be used to help develop new therapeutics and diagnostics. Within the design pipeline, accurate antibody structure modeling is essential, as it provides the basis for antibody-antigen docking, binding affinity prediction, and estimating thermal stability. Ideally, models should be rapidly generated, allowing the exploration of the breadth of antibody space. This allows methods to replicate the natural processes of antibody diversification (e.g., V(D)J recombination and somatic hypermutation), and cope with large volumes of data that are typical of next-generation sequencing datasets. Here we describe ABodyBuilder and PEARS, algorithms that build and mutate antibody model structures. These methods take ~30 s to generate a model antibody structure.

Non-H3 CDR template selection in antibody modeling through machine learning

Antibodies are proteins generated by the adaptive immune system to recognize and counteract a plethora of pathogens through specific binding. This adaptive binding is mediated by structural diversity in the six complementary determining region (CDR) loops (H1, H2, H3, L1, L2 and L3), which also makes accurate structural modeling of CDRs challenging. Both homology and de novo modeling approaches have been used; to date, the former has achieved greater accuracy for the non-H3 loops. The homology modeling of non-H3 CDRs is more accurate because non-H3 CDR loops of the same length and type can be grouped into a few structural clusters. Most antibody-modeling suites utilize homology modeling for the non-H3 CDRs, differing only in the alignment algorithm and how/if they utilize structural clusters. While RosettaAntibody and SAbPred do not explicitly assign query CDR sequences to clusters, two other approaches, PIGS and Kotai Antibody Builder, utilize sequence-based rules to assign CDR sequences to clusters. While the manually curated sequence rules can identify better structural templates, because their curation requires extensive literature search and human effort, they lag behind the deposition of new antibody structures and are infrequently updated. In this study, we propose a machine learning approach (Gradient Boosting Machine [GBM]) to learn the structural clusters of non-H3 CDRs from sequence alone. The GBM method simplifies feature selection and can easily integrate new data, compared to manual sequence rule curation. We compare the classification results using the GBM method to that of RosettaAntibody in a 3-repeat 10-fold cross-validation (CV) scheme on the cluster-annotated antibody database PyIgClassify and we observe an improvement in the classification accuracy of the concerned loops from 84.5% ± 0.24% to 88.16% ± 0.056%. The GBM models reduce the errors in specific cluster membership misclassifications when the involved clusters have relatively abundant data. Based on the factors identified, we suggest methods that can enrich structural classes with sparse data to further improve prediction accuracy in future studies.

Modeling of Protein Tertiary and Quaternary Structures Based on Evolutionary Information

Proteins are subject to evolutionary forces that shape their three-dimensional structure to meet specific functional demands. The knowledge of the structure of a protein is therefore instrumental to gain information about the molecular basis of its function. However, experimental structure determination is inherently time consuming and expensive, making it impossible to follow the explosion of sequence data deriving from genome-scale projects. As a consequence, computational structural modeling techniques have received much attention and established themselves as a valuable complement to experimental structural biology efforts. Among these, comparative modeling remains the method of choice to model the three-dimensional structure of a protein when homology to a protein of known structure can be detected.The general strategy consists of using experimentally determined structures of proteins as templates for the generation of three-dimensional models of related family members (targets) of which the structure is unknown. This chapter provides a description of the individual steps needed to obtain a comparative model using SWISS-MODEL, one of the most widely used automated servers for protein structure homology modeling.

Automated shape-based clustering of 3D immunoglobulin protein structures in chronic lymphocytic leukemia

Background

Although the etiology of chronic lymphocytic leukemia (CLL), the most common type of adult leukemia, is still unclear, strong evidence implicates antigen involvement in disease ontogeny and evolution. Primary and 3D structure analysis has been utilised in order to discover indications of antigenic pressure. The latter has been mostly based on the 3D models of the clonotypic B cell receptor immunoglobulin (BcR IG) amino acid sequences. Therefore, their accuracy is directly dependent on the quality of the model construction algorithms and the specific methods used to compare the ensuing models. Thus far, reliable and robust methods that can group the IG 3D models based on their structural characteristics are missing.

Results

Here we propose a novel method for clustering a set of proteins based on their 3D structure focusing on 3D structures of BcR IG from a large series of patients with CLL. The method combines techniques from the areas of bioinformatics, 3D object recognition and machine learning. The clustering procedure is based on the extraction of 3D descriptors, encoding various properties of the local and global geometrical structure of the proteins. The descriptors are extracted from aligned pairs of proteins. A combination of individual 3D descriptors is also used as an additional method. The comparison of the automatically generated clusters to manual annotation by experts shows an increased accuracy when using the 3D descriptors compared to plain bioinformatics-based comparison. The accuracy is increased even more when using the combination of 3D descriptors.

Conclusions

The experimental results verify that the use of 3D descriptors commonly used for 3D object recognition can be effectively applied to distinguishing structural differences of proteins. The proposed approach can be applied to provide hints for the existence of structural groups in a large set of unannotated BcR IG protein files in both CLL and, by logical extension, other contexts where it is relevant to characterize BcR IG structural similarity. The method does not present any limitations in application and can be extended to other types of proteins.

SWISS-MODEL: homology modelling of protein structures and complexes

Homology modelling has matured into an important technique in structural biology, significantly contributing to narrowing the gap between known protein sequences and experimentally determined structures. Fully automated workflows and servers simplify and streamline the homology modelling process, also allowing users without a specific computational expertise to generate reliable protein models and have easy access to modelling results, their visualization and interpretation. Here, we present an update to the SWISS-MODEL server, which pioneered the field of automated modelling 25 years ago and been continuously further developed. Recently, its functionality has been extended to the modelling of homo- and heteromeric complexes. Starting from the amino acid sequences of the interacting proteins, both the stoichiometry and the overall structure of the complex are inferred by homology modelling. Other major improvements include the implementation of a new modelling engine, ProMod3 and the introduction a new local model quality estimation method, QMEANDisCo. SWISS-MODEL is freely available at https://swissmodel.expasy.org.

Sphinx: merging knowledge-based and ab initio approaches to improve protein loop prediction

Motivation

Loops are often vital for protein function, however, their irregular structures make them difficult to model accurately. Current loop modelling algorithms can mostly be divided into two categories: knowledge-based, where databases of fragments are searched to find suitable conformations and ab initio, where conformations are generated computationally. Existing knowledge-based methods only use fragments that are the same length as the target, even though loops of slightly different lengths may adopt similar conformations. Here, we present a novel method, Sphinx, which combines ab initio techniques with the potential extra structural information contained within loops of a different length to improve structure prediction.

Results

We show that Sphinx is able to generate high-accuracy predictions and decoy sets enriched with near-native loop conformations, performing better than the ab initio algorithm on which it is based. In addition, it is able to provide predictions for every target, unlike some knowledge-based methods. Sphinx can be used successfully for the difficult problem of antibody H3 prediction, outperforming RosettaAntibody, one of the leading H3-specific ab initio methods, both in accuracy and speed.

Availability and Implementation

Sphinx is available at http://opig.stats.ox.ac.uk/webapps/sphinx.

Supplementary information

Supplementary data are available at Bioinformatics online.

Role of Non-local Interactions between CDR Loops in Binding Affinity of MR78 Antibody to Marburg Virus Glycoprotein

An atomic-detail model of the Marburg virus glycoprotein in complex with a neutralizing human monoclonal antibody designated MR78 was constructed using Phenix.Rosetta starting from a 3.6Å crystallographic density map. The Asp at T6 in the HCDR3's bulged torso cannot form the canonical salt bridge as position T2 lacks an Arg or Lys residue. It instead engages in a hydrogen bond interaction with a Tyr contributed by the HCDR1 loop. This inter-CDR loop interaction stabilizes the bulged conformation needed for binding to the viral glycoprotein: a Tyr to Phe mutant displays a binding affinity reduced by a factor of at least 10. We found that 5% of a database of 465 million human antibody sequences has the same residues at T2 and T6 positions in HCDR3 and Tyr in HCDR1 that could potentially form this Asp-Tyr interaction, and that this interaction might contribute to a non-canonical bulged torso conformation.

Superposition-free comparison and clustering of antibody binding sites: implications for the prediction of the nature of their antigen

We describe here a superposition free method for comparing the surfaces of antibody binding sites based on the Zernike moments and show that they can be used to quickly compare and cluster sets of antibodies. The clusters provide information about the nature of the bound antigen that, when combined with a method for predicting the number of direct antibody antigen contacts, allows the discrimination between protein and non-protein binding antibodies with an accuracy of 76%. This is of relevance in several aspects of antibody science, for example to select the framework to be used for a combinatorial antibody library.

Antibody H3 Structure Prediction

Antibodies are proteins of the immune system that are able to bind to a huge variety of different substances, making them attractive candidates for therapeutic applications. Antibody structures have the potential to be useful during drug development, allowing the implementation of rational design procedures. The most challenging part of the antibody structure to experimentally determine or model is the H3 loop, which in addition is often the most important region in an antibody's binding site. This review summarises the approaches used so far in the pursuit of accurate computational H3 structure prediction.

Computer-Aided Antibody Design: An Overview

The use of monoclonal antibody as the next generation protein therapeutics with remarkable success has surged the development of antibody engineering to design molecules for optimizing affinity, better efficacy, greater safety and therapeutic function. Therefore, computational methods have become increasingly important to generate hypotheses, interpret and guide experimental works. In this chapter, we discussed the overall antibody design by computational approches.

Distance-Guided Forward and Backward Chain-Growth Monte Carlo Method for Conformational Sampling and Structural Prediction of Antibody CDR-H3 Loops

Antibodies recognize antigens through the complementary determining regions (CDR) formed by six-loop hypervariable regions crucial for the diversity of antigen specificities. Among the six CDR loops, the H3 loop is the most challenging to predict because of its much higher variation in sequence length and identity, resulting in much larger and complex structural space, compared to the other five loops. We developed a novel method based on a chain-growth sequential Monte Carlo method, called distance-guided sequential chain-growth Monte Carlo for H3 loops (DiSGro-H3). The new method samples protein chains in both forward and backward directions. It can efficiently generate low energy, near-native H3 loop structures using the conformation types predicted from the sequences of H3 loops. DiSGro-H3 performs significantly better than another ab initio method, RosettaAntibody, in both sampling and prediction, while taking less computational time. It performs comparably to template-based methods. As an ab initio method, DiSGro-H3 offers satisfactory accuracy while being able to predict any H3 loops without templates.

LoopIng: a template-based tool for predicting the structure of protein loops

Motivation: Predicting the structure of protein loops is very challenging, mainly because they are not necessarily subject to strong evolutionary pressure. This implies that, unlike the rest of the protein, standard homology modeling techniques are not very effective in modeling their structure. However, loops are often involved in protein function, hence inferring their structure is important for predicting protein structure as well as function.

Results: We describe a method, LoopIng, based on the Random Forest automated learning technique, which, given a target loop, selects a structural template for it from a database of loop candidates. Compared to the most recently available methods, LoopIng is able to achieve similar accuracy for short loops (4–10 residues) and significant enhancements for long loops (11–20 residues). The quality of the predictions is robust to errors that unavoidably affect the stem regions when these are modeled. The method returns a confidence score for the predicted template loops and has the advantage of being very fast (on average: 1 min/loop).

Availability and implementation:www.biocomputing.it/looping

Contact:anna.tramontano@uniroma1.it

Supplementary information:Supplementary data are available at Bioinformatics online.

LYRA, a webserver for lymphocyte receptor structural modeling

The accurate structural modeling of B- and T-cell receptors is fundamental to gain a detailed insight in the mechanisms underlying immunity and in developing new drugs and therapies. The LYRA (LYmphocyte Receptor Automated modeling) web server (http://www.cbs.dtu.dk/services/LYRA/) implements a complete and automated method for building of B- and T-cell receptor structural models starting from their amino acid sequence alone. The webserver is freely available and easy to use for non-specialists. Upon submission, LYRA automatically generates alignments using ad hoc profiles, predicts the structural class of each hypervariable loop, selects the best templates in an automatic fashion, and provides within minutes a complete 3D model that can be downloaded or inspected online. Experienced users can manually select or exclude template structures according to case specific information. LYRA is based on the canonical structure method, that in the last 30 years has been successfully used to generate antibody models of high accuracy, and in our benchmarks this approach proves to achieve similarly good results on TCR modeling, with a benchmarked average RMSD accuracy of 1.29 and 1.48 Å for B- and T-cell receptors, respectively. To the best of our knowledge, LYRA is the first automated server for the prediction of TCR structure.

Antibody modeling using the prediction of immunoglobulin structure (PIGS) web server [corrected]

Antibodies (or immunoglobulins) are crucial for defending organisms from pathogens, but they are also key players in many medical, diagnostic and biotechnological applications. The ability to predict their structure and the specific residues involved in antigen recognition has several useful applications in all of these areas. Over the years, we have developed or collaborated in developing a strategy that enables researchers to predict the 3D structure of antibodies with a very satisfactory accuracy. The strategy is completely automated and extremely fast, requiring only a few minutes (∼10 min on average) to build a structural model of an antibody. It is based on the concept of canonical structures of antibody loops and on our understanding of the way light and heavy chains pack together.