A compact vocabulary of paratope-epitope interactions enables predictability of antibody-antigen binding

Structural trends in antibody-antigen binding interfaces: a computational analysis of 1833 experimentally determined 3D structures

Antibodies are attractive therapeutic candidates due to their ability to bind cognate antigens with high affinity and specificity. Still, the underlying molecular rules governing the antibody-antigen interface remain poorly understood, making in silico antibody design inherently difficult and keeping the discovery and design of novel antibodies a costly and laborious process. This study investigates the characteristics of antibody-antigen binding interfaces through a computational analysis of more than 850,000 atom-atom contacts from the largest reported set of antibody-antigen complexes with 1833 nonredundant, experimentally determined structures. The analysis compares binding characteristics of conventional antibodies and single-domain antibodies (sdAbs) targeting both protein- and peptide antigens. We find clear patterns in the number antibody-antigen contacts and amino acid frequencies in the paratope. The direct comparison of sdAbs and conventional antibodies helps elucidate the mechanisms employed by sdAbs to compensate for their smaller size and the fact that they harbor only half the number of complementarity-determining regions compared to conventional antibodies. Furthermore, we pinpoint antibody interface hotspot residues that are often found at the binding interface and the amino acid frequencies at these positions. These findings have direct potential applications in antibody engineering and the design of improved antibody libraries.

A general approach for selection of epitope-directed binders to proteins

Directing antibodies to a particular epitope among many possible on a target protein is a significant challenge. Here, we present a simple and general method for epitope-directed selection (EDS) using a differential phage selection strategy. This involves engineering the protein of interest (POI) with the epitope of interest (EOI) mutated using a systematic bioinformatics algorithm to guide the local design of an EOI decoy variant. Using several alternating rounds of negative selection with the EOI decoy variant followed by positive selection on the wild-type POI, we were able to identify highly specific and potent antibodies to five different EOI antigens that bind and functionally block known sites of proteolysis. Among these, we developed highly specific antibodies that target the proteolytic site on the CUB domain containing protein 1 (CDCP1) to prevent its proteolysis allowing us to study the cellular maturation of this event that triggers malignancy. We generated antibodies that recognize the junction between the pro- and catalytic domains for three different matrix metalloproteases (MMPs), MMP1, MMP3, and MMP9, that selectively block activation of each of these enzymes and impair cell migration. We targeted a proteolytic epitope on the cell surface receptor, EPH Receptor A2 (EphA2), that is known to transform it from a tumor suppressor to an oncoprotein. We believe that the EDS method greatly facilitates the generation of antibodies to specific EOIs on a wide range of proteins and enzymes for broad therapeutic and diagnostic applications.

Leveraging Cross-Linking Mass Spectrometry for Modeling Antibody-Antigen Complexes

Elucidating antibody-antigen complexes at the atomic level is of utmost interest for understanding immune responses and designing better therapies. Cross-linking mass spectrometry (XL-MS) has emerged as a powerful tool for mapping protein-protein interactions, suggesting valuable structural insights. However, the use of XL-MS studies to enable epitope/paratope mapping of antibody-antigen complexes is still limited up to now. XL-MS data can be used to drive integrative modeling of antibody-antigen complexes, where cross-links information serves as distance restraints for the precise determination of binding interfaces. In this approach, XL-MS data are employed to identify connections between binding interfaces of the antibody and the antigen, thus informing molecular modeling. Current literature provides minimal input about the impact of XL-MS data on the integrative modeling of antibody-antigen complexes. Here, we applied XL-MS to retrieve information about binding interfaces of three antibody-antigen complexes. We leveraged XL-MS data to perform integrative modeling using HADDOCK (active-passive residues and distance restraints strategies) and AlphaLink2. We then compared these three approaches with initial predictions of investigated antibody-antigen complexes by AlphaFold Multimer. This work emphasizes the importance of cross-linking data in resolving conformational dynamics of antibody-antigen complexes, ultimately enhancing the design of better protein therapeutics and vaccines.

Building Representation Learning Models for Antibody Comprehension

Antibodies are versatile proteins with both the capacity to bind a broad range of targets and a proven track record as some of the most successful therapeutics. However, the development of novel antibody therapeutics is a lengthy and costly process. It is challenging to predict the functional and biophysical properties of antibodies from their amino acid sequence alone, requiring numerous experiments for full characterization. Machine learning, specifically deep representation learning, has emerged as a family of methods that can complement wet lab approaches and accelerate the overall discovery and engineering process. Here, we review advances in antibody sequence representation learning, and how this has improved antibody structure prediction and facilitated antibody optimization. We discuss challenges in the development and implementation of such models, such as the lack of publicly available, well-curated antibody function data and highlight opportunities for improvement. These and future advances in machine learning for antibody sequences have the potential to increase the success rate in developing new therapeutics, resulting in broader access to transformative medicines and improved patient outcomes.

Revolutionizing Synthetic Antibody Design: Harnessing Artificial Intelligence and Deep Sequencing Big Data for Unprecedented Advances

Synthetic antibodies (Abs) represent a category of engineered proteins meticulously crafted to replicate the functions of their natural counterparts. Such Abs are generated in vitro, enabling advanced molecular alterations associated with antigen recognition, paratope site engineering, and biochemical refinements. In a parallel realm, deep sequencing has brought about a paradigm shift in molecular biology. It facilitates the prompt and cost-effective high-throughput sequencing of DNA and RNA molecules, enabling the comprehensive big data analysis of Ab transcriptomes, including specific regions of interest. Significantly, the integration of artificial intelligence (AI), based on machine- and deep- learning approaches, has fundamentally transformed our capacity to discern patterns hidden within deep sequencing big data, including distinctive Ab features and protein folding free energy landscapes. Ultimately, current AI advances can generate approximations of the most stable Ab structural configurations, enabling the prediction of de novo synthetic Abs. As a result, this manuscript comprehensively examines the latest and relevant literature concerning the intersection of deep sequencing big data and AI methodologies for the design and development of synthetic Abs. Together, these advancements have accelerated the exploration of antibody repertoires, contributing to the refinement of synthetic Ab engineering and optimizations, and facilitating advancements in the lead identification process.

Next generation of multispecific antibody engineering

Multispecific antibodies recognize two or more epitopes located on the same or distinct targets. This added capability through protein design allows these man-made molecules to address unmet medical needs that are no longer possible with single targeting such as with monoclonal antibodies or cytokines alone. However, the approach to the development of these multispecific molecules has been met with numerous road bumps, which suggests that a new workflow for multispecific molecules is required. The investigation of the molecular basis that mediates the successful assembly of the building blocks into non-native quaternary structures will lead to the writing of a playbook for multispecifics. This is a must do if we are to design workflows that we can control and in turn predict success. Here, we reflect on the current state-of-the-art of therapeutic biologics and look at the building blocks, in terms of proteins, and tools that can be used to build the foundations of such a next-generation workflow.

Convergent evolution and B-cell recirculation in germinal centers in a human lymph node

Germinal centers (GCs) play a central role in generating an effective immune response against infectious pathogens, and failures in their regulating mechanisms can lead to the development of autoimmune diseases and cancer. Although previous works study experimental systems of the immune response with mouse models that are immunized with specific antigens, our study focused on a real-life situation, with an ongoing GC response in a human lymph node (LN) involving multiple asynchronized GCs reacting simultaneously to unknown antigens. We combined laser capture microdissection of individual GCs from human LN with next-generation repertoire sequencing to characterize individual GCs as distinct evolutionary spaces. In line with well-characterized GC responses in mice, elicited by immunization with model antigens, we observe a heterogeneous clonal diversity across individual GCs from the same human LN. Still, we identify shared clones in several individual GCs, and phylogenetic tree analysis combined with paratope modeling suggest the re-engagement and rediversification of B-cell clones across GCs and expanded clones exhibiting shared antigen responses across distinct GCs, indicating convergent evolution of the GCs.

Identifying promising sequences for protein engineering using a deep transformer protein language model

Protein engineers aim to discover and design novel sequences with targeted, desirable properties. Given the near limitless size of the protein sequence landscape, it is no surprise that these desirable sequences are often a relative rarity. This makes identifying such sequences a costly and time-consuming endeavor. In this work, we show how to use a deep transformer protein language model to identify sequences that have the most promise. Specifically, we use the model's self-attention map to calculate a Promise Score that weights the relative importance of a given sequence according to predicted interactions with a specified binding partner. This Promise Score can then be used to identify strong binders worthy of further study and experimentation. We use the Promise Score within two protein engineering contexts-Nanobody (Nb) discovery and protein optimization. With Nb discovery, we show how the Promise Score provides an effective way to select lead sequences from Nb repertoires. With protein optimization, we show how to use the Promise Score to select site-specific mutagenesis experiments that identify a high percentage of improved sequences. In both cases, we also show how the self-attention map used to calculate the Promise Score can indicate which regions of a protein are involved in intermolecular interactions that drive the targeted property. Finally, we describe how to fine-tune the transformer protein language model to learn a predictive model for the targeted property, and discuss the capabilities and limitations of fine-tuning with and without knowledge transfer within the context of protein engineering.

Insights into next generation sequencing guided antibody selection strategies

Therapeutic antibody discovery often relies on in-vitro display methods to identify lead candidates. Assessing selected output diversity traditionally involves random colony picking and Sanger sequencing, which has limitations. Next-generation sequencing (NGS) offers a cost-effective solution with increased read depth, allowing a comprehensive understanding of diversity. Our study establishes NGS guidelines for antibody drug discovery, demonstrating its advantages in expanding the number of unique HCDR3 clusters, broadening the number of high affinity antibodies, expanding the total number of antibodies recognizing different epitopes, and improving lead prioritization. Surprisingly, our investigation into the correlation between NGS-derived frequencies of CDRs and affinity revealed a lack of association, although this limitation could be moderately mitigated by leveraging NGS clustering, enrichment and/or relative abundance across different regions to enhance lead prioritization. This study highlights NGS benefits, offering insights, recommendations, and the most effective approach to leverage NGS in therapeutic antibody discovery.

Statistical Analysis and Tokenization of Epitopes to Construct Artificial Neoepitope Libraries

Epitopes are specific regions on an antigen's surface that the immune system recognizes. Epitopes are usually protein regions on foreign immune-stimulating entities such as viruses and bacteria, and in some cases, endogenous proteins may act as antigens. Identifying epitopes is crucial for accelerating the development of vaccines and immunotherapies. However, mapping epitopes in pathogen proteomes is challenging using conventional methods. Screening artificial neoepitope libraries against antibodies can overcome this issue. Here, we applied conventional sequence analysis and methods inspired in natural language processing to reveal specific sequence patterns in the linear epitopes deposited in the Immune Epitope Database (www.iedb.org) that can serve as building blocks for the design of universal epitope libraries. Our results reveal that amino acid frequency in annotated linear epitopes differs from that in the human proteome. Aromatic residues are overrepresented, while the presence of cysteines is practically null in epitopes. Byte pair encoding tokenization shows high frequencies of tryptophan in tokens of 5, 6, and 7 amino acids, corroborating the findings of the conventional sequence analysis. These results can be applied to reduce the diversity of linear epitope libraries by orders of magnitude.

Assessment of neutralization susceptibility of Omicron subvariants XBB.1.5 and BQ.1.1 against broad-spectrum neutralizing antibodies through epitopes mapping

The emergence of new variants of the SARS-CoV-2 virus has posed a significant challenge in developing broadly neutralizing antibodies (nAbs) with guaranteed therapeutic potential. Some nAbs, such as Sotrovimab, have exhibited varying levels of efficacy against different variants, while others, such as Bebtelovimab and Bamlanivimab-etesevimab are ineffective against specific variants, including BQ.1.1 and XBB. This highlights the urgent need for developing broadly active monoclonal antibodies (mAbs) providing prophylactic and therapeutic benefits to high-risk patients, especially in the face of the risk of reinfection from new variants. Here, we aimed to investigate the feasibility of redirecting existing mAbs against new variants of SARS-CoV-2, as well as to understand how BQ.1.1 and XBB.1.5 can evade broadly neutralizing mAbs. By mapping epitopes and escape sites, we discovered that the new variants evade multiple mAbs, including FDA-approved Bebtelovimab, which showed resilience against other Omicron variants. Our approach, which included simulations, endpoint free energy calculation, and shape complementarity analysis, revealed the possibility of identifying mAbs that are effective against both BQ.1.1 and XBB.1.5. We identified two broad-spectrum mAbs, R200-1F9 and R207-2F11, as potential candidates with increased binding affinity to XBB.1.5 and BQ.1.1 compared to the reference (Wu01) strain. Additionally, we propose that these mAbs do not interfere with Angiotensin Converting Enzyme 2 (ACE2) and bind to conserved epitopes on the receptor binding domain of Spike that are not-overlapping, potentially providing a solution to neutralize these new variants either independently or as part of a combination (cocktail) treatment.

Antibody designing against IIIabc junction (JIIIabc) of HCV IRES through affinity maturation; RNA-Antibody docking and interaction analysis

Hepatitis C virus is a single-stranded RNA based virus which can cause chronic HCV and hepatocellular carcinoma. HCV genotype 3a has relatively higher rate of fibrosis progression, prevalence of steatosis and incidence of HCC. Despite HCVs variation in genomic sequence, the 5' untranslated region containing internal ribosome entry site (IRES) is highly conserved among all genotypes. It is responsible for translation and initiation of the viral protein. In present study, IRES was targeted by designing variants of reported antigen binding fragment (Fab) through affinity maturation approach. Affinity maturation strategy allowed the rational antibody designing with better biophysical properties and antibody-antigen binding interactions. Complementarity determining regions of reported Fab (wild type) were assessed and docked with IRES. Best generated model of Fab was selected and subjected to alanine scanning Three sets of insilico mutations for variants (V) designing were selected; single (1-71), double (a-j) and triple (I-X). Redocking of IRES-Fab variants consequently enabled the discovery of three variants exhibiting better docking score as compared to the wild type Fab. V1, V39 and V4 exhibited docking scores of -446.51, -446.52 and-446.29 kcal/mol respectively which is better as compared to the wild type Fab that exhibited the docking score of -351.23 kcal/mol. Variants exhibiting better docking score were screened for aggregation propensity by assessing the aggregation prone regions in Fab structure. Total A3D scores of wild type Fab, V1, V4 and V39 were predicted as -315.325, -312.727, -316.967 and -317.545 respectively. It is manifested that solubility of V4 and V39 is comparable to wild type Fab. In future, development and invitro assessment of these promising Fab HCV3 variants is aimed.

Germinal center-dependent and -independent immune responses of tumor-infiltrating B cells in human cancers

B cells play essential roles in immunity, mainly through the production of high affinity plasma cells (PCs) and memory B (Bmem) cells. The affinity maturation and differentiation of B cells rely on the integration of B-cell receptor (BCR) intrinsic and extrinsic signals provided by antigen binding and the microenvironment, respectively. In recent years, tumor infiltrating B (TIL-B) cells and PCs (TIL-PCs) have been revealed as important players in antitumor responses in human cancers, but their interplay and dynamics remain largely unknown. In lymphoid organs, B-cell responses involve both germinal center (GC)-dependent and GC-independent pathways for Bmem cell and PC production. Affinity maturation of BCR repertoires occurs in GC reactions with specific spatiotemporal dynamics of signal integration by B cells. In general, the reactivation of high-affinity Bmem cells by antigens triggers GC-independent production of large numbers of PC without BCR rediversification. Understanding B-cell dynamics in immune responses requires the integration of multiple tools and readouts such as single-cell phenotyping and RNA-seq, in situ analyses, BCR repertoire analysis, BCR specificity and affinity assays, and functional tests. Here, we review how those tools have recently been applied to study TIL-B cells and TIL-PC in different types of solid tumors. We assessed the published evidence for different models of TIL-B-cell dynamics involving GC-dependent or GC-independent local responses and the resulting production of antigen-specific PCs. Altogether, we highlight the need for more integrative B-cell immunology studies to rationally investigate TIL-B cells as a leverage for antitumor therapies.

Perspectives of Proteomics in Respiratory Allergic Diseases

Proteomics in respiratory allergic diseases has such a battery of techniques and programs that one would almost think there is nothing impossible to find, invent or mold. All the resources that we document here are involved in solving problems in allergic diseases, both diagnostic and prognostic treatment, and immunotherapy development. The main perspectives, according to this version, are in three strands and/or a lockout immunological system: (1) Blocking the diapedesis of the cells involved, (2) Modifications and blocking of paratopes and epitopes being understood by modifications to antibodies, antagonisms, or blocking them, and (3) Blocking FcεRI high-affinity receptors to prevent specific IgEs from sticking to mast cells and basophils. These tools and targets in the allergic landscape are, in our view, the prospects in the field. However, there are still many allergens to identify, including some homologies between allergens and cross-reactions, through the identification of structures and epitopes. The current vision of using proteomics for this purpose remains a constant; this is also true for the basis of diagnostic and controlled systems for immunotherapy. Ours is an open proposal to use this vision for treatment.

Recent Progress in Antibody Epitope Prediction

Recent progress in epitope prediction has shown promising results in the development of vaccines and therapeutics against various diseases. However, the overall accuracy and success rate need to be improved greatly to gain practical application significance, especially conformational epitope prediction. In this review, we examined the general features of antibody-antigen recognition, highlighting the conformation selection mechanism in flexible antibody-antigen binding. We recently highlighted the success and warning signs of antibody epitope predictions, including linear and conformation epitope predictions. While deep learning-based models gradually outperform traditional feature-based machine learning, sequence and structure features still provide insight into antibody-antigen recognition problems.

How can we discover developable antibody-based biotherapeutics?

Antibody-based biotherapeutics have emerged as a successful class of pharmaceuticals despite significant challenges and risks to their discovery and development. This review discusses the most frequently encountered hurdles in the research and development (R&D) of antibody-based biotherapeutics and proposes a conceptual framework called biopharmaceutical informatics. Our vision advocates for the syncretic use of computation and experimentation at every stage of biologic drug discovery, considering developability (manufacturability, safety, efficacy, and pharmacology) of potential drug candidates from the earliest stages of the drug discovery phase. The computational advances in recent years allow for more precise formulation of disease concepts, rapid identification, and validation of targets suitable for therapeutic intervention and discovery of potential biotherapeutics that can agonize or antagonize them. Furthermore, computational methods for de novo and epitope-specific antibody design are increasingly being developed, opening novel computationally driven opportunities for biologic drug discovery. Here, we review the opportunities and limitations of emerging computational approaches for optimizing antigens to generate robust immune responses, in silico generation of antibody sequences, discovery of potential antibody binders through virtual screening, assessment of hits, identification of lead drug candidates and their affinity maturation, and optimization for developability. The adoption of biopharmaceutical informatics across all aspects of drug discovery and development cycles should help bring affordable and effective biotherapeutics to patients more quickly.

Defining and Studying B Cell Receptor and TCR Interactions

BCRs (Abs) and TCRs (or adaptive immune receptors [AIRs]) are the means by which the adaptive immune system recognizes foreign and self-antigens, playing an integral part in host defense, as well as the emergence of autoimmunity. Importantly, the interaction between AIRs and their cognate Ags defies a simple key-in-lock paradigm and is instead a complex many-to-many mapping between an individual's massively diverse AIR repertoire, and a similarly diverse antigenic space. Understanding how adaptive immunity balances specificity with epitopic coverage is a key challenge for the field, and terms such as broad specificity, cross-reactivity, and polyreactivity remain ill-defined and are used inconsistently. In this Immunology Notes and Resources article, a group of experimental, structural, and computational immunologists define commonly used terms associated with AIR binding, describe methodologies to study these binding modes, as well as highlight the implications of these different binding modes for therapeutic design.

Efficient encoding of large antigenic spaces by epitope prioritization with Dolphyn

We investigated a relatively underexplored component of the gut-immune axis by profiling the antibody response to gut phages using Phage Immunoprecipitation Sequencing (PhIP-Seq). To enhance this approach, we developed Dolphyn, a novel method that uses machine learning to select peptides from protein sets and compresses the proteome through epitope-stitching. Dolphyn improves the fraction of gut phage library peptides bound by antibodies from 10% to 31% in healthy individuals, while also reducing the number of synthesized peptides by 78%. In our study on gut phages, we discovered that the immune system develops antibodies to bacteria-infecting viruses in the human gut, particularly E.coli-infecting Myoviridae. Cost-effective PhIP-Seq libraries designed with Dolphyn enable the assessment of a wider range of proteins in a single experiment, thus facilitating the study of the gut-immune axis.

Specific attributes of the VL domain influence both the structure and structural variability of CDR-H3 through steric effects

Antibodies, through their ability to target virtually any epitope, play a key role in driving the adaptive immune response in jawed vertebrates. The binding domains of standard antibodies are their variable light (VL) and heavy (VH) domains, both of which present analogous complementarity-determining region (CDR) loops. It has long been known that the VH CDRs contribute more heavily to the antigen-binding surface (paratope), with the CDR-H3 loop providing a major modality for the generation of diverse paratopes. Here, we provide evidence for an additional role of the VL domain as a modulator of CDR-H3 structure, using a diverse set of antibody crystal structures and a large set of molecular dynamics simulations. We show that specific attributes of the VL domain such as subtypes, CDR canonical forms and genes can influence the structural diversity of the CDR-H3 loop, and provide a physical model for how this effect occurs through inter-loop contacts and packing of CDRs against each other. Our results indicate that the rigid minor loops fine-tune the structure of CDR-H3, thereby contributing to the generation of surfaces complementary to the vast number of possible epitope topologies, and provide insights into the interdependent nature of CDR conformations, an understanding of which is important for the rational antibody design process.

The ClusPro AbEMap web server for the prediction of antibody epitopes

Antibodies play an important role in the immune system by binding to molecules called antigens at their respective epitopes. These interfaces or epitopes are structural entities determined by the interactions between an antibody and an antigen, making them ideal systems to analyze by using docking programs. Since the advent of high-throughput antibody sequencing, the ability to perform epitope mapping using only the sequence of the antibody has become a high priority. ClusPro, a leading protein-protein docking server, together with its template-based modeling version, ClusPro-TBM, have been re-purposed to map epitopes for specific antibody-antigen interactions by using the Antibody Epitope Mapping server (AbEMap). ClusPro-AbEMap offers three different modes for users depending on the information available on the antibody as follows: (i) X-ray structure, (ii) computational/predicted model of the structure or (iii) only the amino acid sequence. The AbEMap server presents a likelihood score for each antigen residue of being part of the epitope. We provide detailed information on the server's capabilities for the three options and discuss how to obtain the best results. In light of the recent introduction of AlphaFold2 (AF2), we also show how one of the modes allows users to use their AF2-generated antibody models as input. The protocol describes the relative advantages of the server compared to other epitope-mapping tools, its limitations and potential areas of improvement. The server may take 45-90 min depending on the size of the proteins.

Untangling immunoglobulin genotype-function associations

Immunoglobulin (IG) genes, encoding B cell receptors (BCRs), are fundamental components of the mammalian immune system, which evolved to recognize the diverse antigenic universe present in nature. To handle these myriad inputs, BCRs are generated through combinatorial recombination of a set of highly polymorphic germline genes, resulting in a vast repertoire of antigen receptors that initiate responses to pathogens and regulate commensals. Following antigen recognition and B cell activation, memory B cells and plasma cells form, allowing for the development of anamnestic antibody (Ab) responses. How inherited variation in IG genes impacts host traits, disease susceptibility, and Ab recall responses is a topic of great interest. Here, we consider approaches to translate emerging knowledge about IG genetic diversity and expressed repertoires to inform our understanding of Ab function in health and disease etiology. As our understanding of IG genetics grows, so will our need for tools to decipher preferences for IG gene or allele usage in different contexts, to better understand antibody responses at the population level.

DeepBindPPI: Protein-Protein Binding Site Prediction Using Attention Based Graph Convolutional Network

Due to the importance of protein-protein interactions in defence mechanism of living body, attempts were made to investigate its attributes, including, but not limited to, binding affinity, and binding region. Contemporary strategies for binding site prediction largely resort to deep learning techniques but turned out to be low precision models. As laboratory experiments for drug discovery tasks utilize this information, increased false positives devalue the computational methods. This emphasize the need to develop enhanced strategies. DeepBindPPI employs deep learning technique to predict the binding regions of proteins, particularly antigen-antibody interaction sites. The results obtained are applied in a docking environment to confirm their correctness. An integration of graph convolutional network with attention mechanism predicts interacting amino acids with improved precision. The model learns the determining factors in interaction from a general pool of proteins and is then fine-tuned using antigen-antibody data. Comparison of the proposed method with existing techniques shows that the developed model has comparable performance. The use of a separate spatial network clearly improved the precision of the proposed method from 0.4 to 0.5. An attempt to utilize the interface information for docking using the HDOCK server gives promising results, with high-quality structures appearing in the top10 ranks.

Leveraging deep learning to improve vaccine design

Deep learning has led to incredible breakthroughs in areas of research, from self-driving vehicles to solutions, to formal mathematical proofs. In the biomedical sciences, however, the revolutionary results seen in other fields are only now beginning to be realized. Vaccine research and development efforts represent an application with high public health significance. Protein structure prediction, immune repertoire analysis, and phylogenetics are three principal areas in which deep learning is poised to provide key advances. Here, we opine on some of the current challenges with deep learning and how they are being addressed. Despite the nascent stage of deep learning applications in immunological studies, there is ample opportunity to utilize this new technology to address the most challenging and burdensome infectious diseases confronting global populations.

Exploring the impact of clonal definition on B-cell diversity: implications for the analysis of immune repertoires

The adaptive immune system has the extraordinary ability to produce a broad range of immunoglobulins that can bind a wide variety of antigens. During adaptive immune responses, activated B cells duplicate and undergo somatic hypermutation in their B-cell receptor (BCR) genes, resulting in clonal families of diversified B cells that can be related back to a common ancestor. Advances in high-throughput sequencing technologies have enabled the high-throughput characterization of B-cell repertoires, however, the accurate identification of clonally related BCR sequences remains a major challenge. In this study, we compare three different clone identification methods on both simulated and experimental data, and investigate their impact on the characterization of B-cell diversity. We observe that different methods lead to different clonal definitions, which affects the quantification of clonal diversity in repertoire data. Our analyses show that direct comparisons between clonal clusterings and clonal diversity of different repertoires should be avoided if different clone identification methods were used to define the clones. Despite this variability, the diversity indices inferred from the repertoires' clonal characterization across samples show similar patterns of variation regardless of the clonal identification method used. We find the Shannon entropy to be the most robust in terms of the variability of diversity rank across samples. Our analysis also suggests that the traditional germline gene alignment-based method for clonal identification remains the most accurate when the complete information about the sequence is known, but that alignment-free methods may be preferred for shorter sequencing read lengths. We make our implementation freely available as a Python library cdiversity.

Germline-encoded amino acid-binding motifs drive immunodominant public antibody responses

Despite the vast diversity of the antibody repertoire, infected individuals often mount antibody responses to precisely the same epitopes within antigens. The immunological mechanisms underpinning this phenomenon remain unknown. By mapping 376 immunodominant "public epitopes" at high resolution and characterizing several of their cognate antibodies, we concluded that germline-encoded sequences in antibodies drive recurrent recognition. Systematic analysis of antibody-antigen structures uncovered 18 human and 21 partially overlapping mouse germline-encoded amino acid-binding (GRAB) motifs within heavy and light V gene segments that in case studies proved critical for public epitope recognition. GRAB motifs represent a fundamental component of the immune system's architecture that promotes recognition of pathogens and leads to species-specific public antibody responses that can exert selective pressure on pathogens.

Alloreactivity and autoreactivity converge to support B cell epitope targeting in transplant rejection

Antibody (Ab) responses against human leukocyte antigen (HLA) proteins mismatched between donor and recipient are leading cause of allograft loss in kidney transplantation. However, therapies targeting alloreactive B cell and Ab-secreting cell (ASC) are lacking, motivating the need to understand how to prevent and abrogate these alloresponses. Using molecular, structural, and proteomic techniques, we profiled the B cell response in a kidney transplant recipient with antibody-mediated rejection and graft loss. We found that this response spanned the rejected organ and peripheral blood, stimulated the differentiation of multiple B cell subsets, and produced a high-affinity, donor-specific, anti-HLA response. We found epitopic immunodominance that relied on highly exposed, solvent-accessible mismatched HLA residues as well as structural and biomolecular evidence of autoreactivity against the recipient's self-HLA allele. These alloreactive and autoreactive signatures converged in the recipient's circulating donor-specific Ab repertoire, suggesting that rejection requires both the recognition of non-self and breaches of tolerance to lead to alloinjury and graft loss.

A specific anti-cyclin D1 intrabody represses breast cancer cell proliferation by interrupting the cyclin D1-CDK4 interaction

BACKGROUND

Cyclin D1 overexpression may contribute to development of various cancers, including breast cancer, and thus may serve as a key cancer diagnostic marker and therapeutic target. In our previous study, we generated a cyclin D1-specific single-chain variable fragment antibody (ADκ) from a human semi-synthetic single-chain variable fragment library. ADκ specifically interacted with recombinant and endogenous cyclin D1 proteins through an unknown molecular basis to inhibit HepG2 cell growth and proliferation.

RESULTS

Here, using phage display and in silico protein structure modeling methods combined with cyclin D1 mutational analysis, key residues that bind to ADκ were identified. Notably, residue K112 within the cyclin box was required for cyclin D1-ADκ binding. In order to elucidate the molecular mechanism underlying ADκ anti-tumor effects, a cyclin D1-specific nuclear localization signal-containing intrabody (NLS-ADκ) was constructed. When expressed within cells, NLS-ADκ interacted specifically with cyclin D1 to significantly inhibit cell proliferation, induce G1-phase arrest, and trigger apoptosis of MCF-7 and MDA-MB-231 breast cancer cells. Moreover, the NLS-ADκ-cyclin D1 interaction blocked binding of cyclin D1 to CDK4 and inhibited RB protein phosphorylation, resulting in altered expression of downstream cell proliferation-related target genes.

CONCLUSION

We identified amino acid residues in cyclin D1 that may play key roles in the ADκ-cyclin D1 interaction. A nuclear localization antibody against cyclin D1 (NLS-ADκ) was constructed and successfully expressed in breast cancer cells. NLS-ADκ exerted tumor suppressor effects via blocking the binding of CDK4 to cyclin D1 and inhibiting phosphorylation of RB. The results presented here demonstrate anti-tumor potential of intrabody-based cyclin D1-targeted breast cancer therapy.

Antibodies to repeat-containing antigens in Plasmodium falciparum are exposure-dependent and short-lived in children in natural malaria infections

Protection against Plasmodium falciparum, which is primarily antibody-mediated, requires recurrent exposure to develop. The study of both naturally acquired limited immunity and vaccine induced protection against malaria remains critical for ongoing eradication efforts. Towards this goal, we deployed a customized P. falciparum PhIP-seq T7 phage display library containing 238,068 tiled 62-amino acid peptides, covering all known coding regions, including antigenic variants, to systematically profile antibody targets in 198 Ugandan children and adults from high and moderate transmission settings. Repeat elements - short amino acid sequences repeated within a protein - were significantly enriched in antibody targets. While breadth of responses to repeat-containing peptides was twofold higher in children living in the high versus moderate exposure setting, no such differences were observed for peptides without repeats, suggesting that antibody responses to repeat-containing regions may be more exposure dependent and/or less durable in children than responses to regions without repeats. Additionally, short motifs associated with seroreactivity were extensively shared among hundreds of antigens, potentially representing cross-reactive epitopes. PfEMP1 shared motifs with the greatest number of other antigens, partly driven by the diversity of PfEMP1 sequences. These data suggest that the large number of repeat elements and potential cross-reactive epitopes found within antigenic regions of P. falciparum could contribute to the inefficient nature of malaria immunity.

Mapping of antibody epitopes based on docking and homology modeling

Antibodies are key proteins produced by the immune system to target pathogen proteins termed antigens via specific binding to surface regions called epitopes. Given an antigen and the sequence of an antibody the knowledge of the epitope is critical for the discovery and development of antibody based therapeutics. In this work, we present a computational protocol that uses template-based modeling and docking to predict epitope residues. This protocol is implemented in three major steps. First, a template-based modeling approach is used to build the antibody structures. We tested several options, including generation of models using AlphaFold2. Second, each antibody model is docked to the antigen using the fast Fourier transform (FFT) based docking program PIPER. Attention is given to optimally selecting the docking energy parameters depending on the input data. In particular, the van der Waals energy terms are reduced for modeled antibodies relative to x-ray structures. Finally, ranking of antigen surface residues is produced. The ranking relies on the docking results, that is, how often the residue appears in the docking poses' interface, and also on the energy favorability of the docking pose in question. The method, called PIPER-Map, has been tested on a widely used antibody-antigen docking benchmark. The results show that PIPER-Map improves upon the existing epitope prediction methods. An interesting observation is that epitope prediction accuracy starting from antibody sequence alone does not significantly differ from that of starting from unbound (i.e., separately crystallized) antibody structure.

Toward real-world automated antibody design with combinatorial Bayesian optimization

Antibodies are multimeric proteins capable of highly specific molecular recognition. The complementarity determining region 3 of the antibody variable heavy chain (CDRH3) often dominates antigen-binding specificity. Hence, it is a priority to design optimal antigen-specific CDRH3 to develop therapeutic antibodies. The combinatorial structure of CDRH3 sequences makes it impossible to query binding-affinity oracles exhaustively. Moreover, antibodies are expected to have high target specificity and developability. Here, we present AntBO, a combinatorial Bayesian optimization framework utilizing a CDRH3 trust region for an in silico design of antibodies with favorable developability scores. The in silico experiments on 159 antigens demonstrate that AntBO is a step toward practically viable in vitro antibody design. In under 200 calls to the oracle, AntBO suggests antibodies outperforming the best binding sequence from 6.9 million experimentally obtained CDRH3s. Additionally, AntBO finds very-high-affinity CDRH3 in only 38 protein designs while requiring no domain knowledge.

The evolutionary and functional significance of germline immunoglobulin gene variation

The recombination between immunoglobulin (IG) gene segments determines an individual's naïve antibody repertoire and, consequently, (auto)antigen recognition. Emerging evidence suggests that mammalian IG germline variation impacts humoral immune responses associated with vaccination, infection, and autoimmunity - from the molecular level of epitope specificity, up to profound changes in the architecture of antibody repertoires. These links between IG germline variants and immunophenotype raise the question on the evolutionary causes and consequences of diversity within IG loci. We discuss why the extreme diversity in IG loci remains a mystery, why resolving this is important for the design of more effective vaccines and therapeutics, and how recent evidence from multiple lines of inquiry may help us do so.

Peptide Microarrays for Studying Autoantibodies in Neurological Disease

Antibody-mediated neurological diseases constitute an emerging clinical entity that remains to be fully explored. Recent studies identified autoantibodies that directly confer pathogenicity, and it was shown that in these cases immunotherapies can result in profound positive patient responses. These advances highlight the urgent need for improved means to effectively screen patient samples for novel autoantibodies (aAbs) and their subsequent characterization. Here, we discuss challenges and opportunities for peptide microarrays to contribute to the identification, mapping, and characterization of the underlying monospecific disease-defining binding surfaces. We outline control experiments, workflow modifications and bioinformatic filtering methods that enhance the predictive power of array-based studies. Further, we highlight experimental and computer-based display approaches that have the potential to expand the use of synthetic microarrays over the detection of discontinuous epitopes. Knowledge over the autoantibody epitopes in neurological disease will enhance our understanding of the pathological mechanisms and thereby potentially contribute to novel diagnostic approaches or even innovative antigen-specific treatments that avoid the serious adverse effects seen with currently used immunosuppressive therapies.

A window into the human immune system: comprehensive characterization of the complexity of antibody complementary-determining regions in functional antibodies

The human immune system uses antibodies to neutralize foreign antigens. They are composed of heavy and light chains, both with constant and variable regions. The variable region has six hypervariable loops, also known as complementary-determining regions (CDRs) that determine antibody diversity and antigen specificity. Knowledge of their significance, and certain residues present in these areas, is vital for antibody therapeutics development. This study includes an analysis of more than 11,000 human antibody sequences from the International Immunogenetics information system (IMGT). The analysis included parameters such as length distribution, overall amino acid diversity, amino acid frequency per CDR and residue position within antibody chains. Overall, our findings confirm existing knowledge, such as CDRH3's high length diversity and amino acid variability, increased aromatic residue usage, particularly tyrosine, charged and polar residues like aspartic acid, serine, and the flexible residue glycine. Specific residue positions within each CDR influence these occurrences, implying a unique amino acid type distribution pattern. We compared amino acid type usage in CDRs and non-CDR regions, both in globular and transmembrane proteins, which revealed distinguishing features, such as increased frequency of tyrosine, serine, aspartic acid, and arginine. These findings should prove useful for future optimization, improvement of affinity, synthetic antibody library design, or the creation of antibodies de-novo in silico.

Toward generalizable prediction of antibody thermostability using machine learning on sequence and structure features

Over the last three decades, the appeal for monoclonal antibodies (mAbs) as therapeutics has been steadily increasing as evident with FDA's recent landmark approval of the 100th mAb. Unlike mAbs that bind to single targets, multispecific biologics (msAbs) have garnered particular interest owing to the advantage of engaging distinct targets. One important modular component of msAbs is the single-chain variable fragment (scFv). Despite the exquisite specificity and affinity of these scFv modules, their relatively poor thermostability often hampers their development as a potential therapeutic drug. In recent years, engineering antibody sequences to enhance their stability by mutations has gained considerable momentum. As experimental methods for antibody engineering are time-intensive, laborious and expensive, computational methods serve as a fast and inexpensive alternative to conventional routes. In this work, we show two machine learning approaches - one with pre-trained language models (PTLM) capturing functional effects of sequence variation, and second, a supervised convolutional neural network (CNN) trained with Rosetta energetic features - to better classify thermostable scFv variants from sequence. Both of these models are trained over temperature-specific data (TS50 measurements) derived from multiple libraries of scFv sequences. On out-of-distribution (refers to the fact that the out-of-distribution sequnes are blind to the algorithm) sequences, we show that a sufficiently simple CNN model performs better than general pre-trained language models trained on diverse protein sequences (average Spearman correlation coefficient, ρ , of 0.4 as opposed to 0.15). On the other hand, an antibody-specific language model performs comparatively better than the CNN model on the same task (ρ = 0.52). Further, we demonstrate that for an independent mAb with available thermal melting temperatures for 20 experimentally characterized thermostable mutations, these models trained on TS50 data could identify 18 residue positions and 5 identical amino-acid mutations showing remarkable generalizability. Our results suggest that such models can be broadly applicable for improving the biological characteristics of antibodies. Further, transferring such models for alternative physicochemical properties of scFvs can have potential applications in optimizing large-scale production and delivery of mAbs or bsAbs.

simAIRR: simulation of adaptive immune repertoires with realistic receptor sequence sharing for benchmarking of immune state prediction methods

BACKGROUND

Machine learning (ML) has gained significant attention for classifying immune states in adaptive immune receptor repertoires (AIRRs) to support the advancement of immunodiagnostics and therapeutics. Simulated data are crucial for the rigorous benchmarking of AIRR-ML methods. Existing approaches to generating synthetic benchmarking datasets result in the generation of naive repertoires missing the key feature of many shared receptor sequences (selected for common antigens) found in antigen-experienced repertoires.

RESULTS

We demonstrate that a common approach to generating simulated AIRR benchmark datasets can introduce biases, which may be exploited for undesired shortcut learning by certain ML methods. To mitigate undesirable access to true signals in simulated AIRR datasets, we devised a simulation strategy (simAIRR) that constructs antigen-experienced-like repertoires with a realistic overlap of receptor sequences. simAIRR can be used for constructing AIRR-level benchmarks based on a range of assumptions (or experimental data sources) for what constitutes receptor-level immune signals. This includes the possibility of making or not making any prior assumptions regarding the similarity or commonality of immune state-associated sequences that will be used as true signals. We demonstrate the real-world realism of our proposed simulation approach by showing that basic ML strategies perform similarly on simAIRR-generated and real-world experimental AIRR datasets.

CONCLUSIONS

This study sheds light on the potential shortcut learning opportunities for ML methods that can arise with the state-of-the-art way of simulating AIRR datasets. simAIRR is available as a Python package: https://github.com/KanduriC/simAIRR.

Switching Heavy Chain Constant Domains Denatures the Paratope 3D Architecture of Influenza Monoclonal Antibodies

Several human monoclonal Abs for treating Influenza have been evaluated in clinical trials with limited success despite demonstrating superiority in preclinical animal models including mice. To conduct efficacy studies in mice, human monoclonal Abs are genetically engineered to contain mouse heavy chain constant domain to facilitate the engagement of Fc-receptors on mouse immune effector cells. Although studies have consistently reported discrepancies in Ab effectiveness following genetic engineering, the structural and mechanistic basis for these inconsistencies remain uncharacterized. Here, we use homology modeling to predict variable region (VR) analogous monoclonal Abs possessing human IgG1, mouse IgG1, and mouse IgG2a heavy chain constant domains. We then examine predicted 3D structures for variations in the spatial location and orientation of corresponding paratope amino acid residues. By structurally aligning crystal structures of Fabs in complex with hemagglutinin (HA), we show that corresponding paratope amino acid residues for VR-analogous human IgG1, mouse IgG1, and mouse IgG2a monoclonal Abs interact differentially with HA suggesting that their epitopes might not be identical. To demonstrate that variations in the paratope 3D fine architecture have implications for Ab specificity and effectiveness, we genetically engineered VR-analogous human IgG1, human IgG4, mouse IgG1, and mouse IgG2a monoclonal Abs and explored their specificity and effectiveness in protecting MDCK cells from infection by pandemic H1N1 and H3N2 Influenza viruses. We found that VR-analogous monoclonal Abs placed on mouse heavy chain constant domains were more efficacious at protecting MDCK cells from Influenza virus infection relative to those on human heavy chain constant domains. Interestingly, mouse but not human heavy chain constant domains increased target breadth in some monoclonal Abs. These data suggest that heavy chain constant domain sequences play a role in shaping Ab repertoires that go beyond class or sub-class differences in immune effector recruitment. This represents a facet of Ab biology that can potentially be exploited to improve the scope and utilization of current therapeutic or prophylactic candidates for influenza.

Unconstrained generation of synthetic antibody-antigen structures to guide machine learning methodology for antibody specificity prediction

Machine learning (ML) is a key technology for accurate prediction of antibody-antigen binding. Two orthogonal problems hinder the application of ML to antibody-specificity prediction and the benchmarking thereof: the lack of a unified ML formalization of immunological antibody-specificity prediction problems and the unavailability of large-scale synthetic datasets to benchmark real-world relevant ML methods and dataset design. Here we developed the Absolut! software suite that enables parameter-based unconstrained generation of synthetic lattice-based three-dimensional antibody-antigen-binding structures with ground-truth access to conformational paratope, epitope and affinity. We formalized common immunological antibody-specificity prediction problems as ML tasks and confirmed that for both sequence- and structure-based tasks, accuracy-based rankings of ML methods trained on experimental data hold for ML methods trained on Absolut!-generated data. The Absolut! framework has the potential to enable real-world relevant development and benchmarking of ML strategies for biotherapeutics design.

Development of therapeutic antibodies for the treatment of diseases

Since the first monoclonal antibody drug, muromonab-CD3, was approved for marketing in 1986, 165 antibody drugs have been approved or are under regulatory review worldwide. With the approval of new drugs for treating a wide range of diseases, including cancer and autoimmune and metabolic disorders, the therapeutic antibody drug market has experienced explosive growth. Monoclonal antibodies have been sought after by many biopharmaceutical companies and scientific research institutes due to their high specificity, strong targeting abilities, low toxicity, side effects, and high development success rate. The related industries and markets are growing rapidly, and therapeutic antibodies are one of the most important research and development areas in the field of biology and medicine. In recent years, great progress has been made in the key technologies and theoretical innovations provided by therapeutic antibodies, including antibody-drug conjugates, antibody-conjugated nuclides, bispecific antibodies, nanobodies, and other antibody analogs. Additionally, therapeutic antibodies can be combined with technologies used in other fields to create new cross-fields, such as chimeric antigen receptor T cells (CAR-T), CAR-natural killer cells (CAR-NK), and other cell therapy. This review summarizes the latest approved or in regulatory review therapeutic antibodies that have been approved or that are under regulatory review worldwide, as well as clinical research on these approaches and their development, and outlines antibody discovery strategies that have emerged during the development of therapeutic antibodies, such as hybridoma technology, phage display, preparation of fully human antibody from transgenic mice, single B-cell antibody technology, and artificial intelligence-assisted antibody discovery.

Differential performance of RoseTTAFold in antibody modeling

Antibodies are essential to life, and knowing their structures can facilitate the understanding of antibody-antigen recognition mechanisms. Precise antibody structure prediction has been a core challenge for a prolonged period, especially the accuracy of H3 loop prediction. Despite recent progress, existing methods cannot achieve atomic accuracy, especially when the homologous structures required for these methods are not available. Recently, RoseTTAFold, a deep learning-based algorithm, has shown remarkable breakthroughs in predicting the 3D structures of proteins. To assess the antibody modeling ability of RoseTTAFold, we first retrieved the sequences of 30 antibodies as the test set and used RoseTTAFold to model their 3D structures. We then compared the models constructed by RoseTTAFold with those of SWISS-MODEL in a different way, in which we stratified Global Model Quality Estimate (GMQE) into three different ranges. The results indicated that RoseTTAFold could achieve results similar to SWISS-MODEL in modeling most CDR loops, especially the templates with a GMQE score under 0.8. In addition, we also compared the structures modeled by RoseTTAFold, SWISS-MODEL and ABodyBuilder. In brief, RoseTTAFold could accurately predict 3D structures of antibodies, but its accuracy was not as good as the other two methods. However, RoseTTAFold exhibited better accuracy for modeling H3 loop than ABodyBuilder and was comparable to SWISS-MODEL. Finally, we discussed the limitations and potential improvements of the current RoseTTAFold, which may help to further the accuracy of RoseTTAFold's antibody modeling.

Machine Learning Approaches to TCR Repertoire Analysis

Sparked by the development of genome sequencing technology, the quantity and quality of data handled in immunological research have been changing dramatically. Various data and database platforms are now driving the rapid progress of machine learning for immunological data analysis. Of various topics in immunology, T cell receptor repertoire analysis is one of the most important targets of machine learning for assessing the state and abnormalities of immune systems. In this paper, we review recent repertoire analysis methods based on machine learning and deep learning and discuss their prospects.

Machine-designed biotherapeutics: opportunities, feasibility and advantages of deep learning in computational antibody discovery

Antibodies are versatile molecular binders with an established and growing role as therapeutics. Computational approaches to developing and designing these molecules are being increasingly used to complement traditional lab-based processes. Nowadays, in silico methods fill multiple elements of the discovery stage, such as characterizing antibody–antigen interactions and identifying developability liabilities. Recently, computational methods tackling such problems have begun to follow machine learning paradigms, in many cases deep learning specifically. This paradigm shift offers improvements in established areas such as structure or binding prediction and opens up new possibilities such as language-based modeling of antibody repertoires or machine-learning-based generation of novel sequences. In this review, we critically examine the recent developments in (deep) machine learning approaches to therapeutic antibody design with implications for fully computational antibody design.

Understanding the mutational frequency in SARS-CoV-2 proteome using structural features

The prolonged transmission of the severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) virus in the human population has led to demographic divergence and the emergence of several location-specific clusters of viral strains. Although the effect of mutation(s) on severity and survival of the virus is still unclear, it is evident that certain sites in the viral proteome are more/less prone to mutations. In fact, millions of SARS-CoV-2 sequences collected all over the world have provided us a unique opportunity to understand viral protein mutations and develop novel computational approaches to predict mutational patterns. In this study, we have classified the mutation sites into low and high mutability classes based on viral isolates count containing mutations. The physicochemical features and structural analysis of the SARS-CoV-2 proteins showed that features including residue type, surface accessibility, residue bulkiness, stability and sequence conservation at the mutation site were able to classify the low and high mutability sites. We further developed machine learning models using above-mentioned features, to predict low and high mutability sites at different selection thresholds (ranging 5-30% of topmost and bottommost mutated sites) and observed the improvement in performance as the selection threshold is reduced (prediction accuracy ranging from 65 to 77%). The analysis will be useful for early detection of variants of concern for the SARS-CoV-2, which can also be applied to other existing and emerging viruses for another pandemic prevention.

Profiling the baseline performance and limits of machine learning models for adaptive immune receptor repertoire classification

BACKGROUND

Machine learning (ML) methodology development for the classification of immune states in adaptive immune receptor repertoires (AIRRs) has seen a recent surge of interest. However, so far, there does not exist a systematic evaluation of scenarios where classical ML methods (such as penalized logistic regression) already perform adequately for AIRR classification. This hinders investigative reorientation to those scenarios where method development of more sophisticated ML approaches may be required.

RESULTS

To identify those scenarios where a baseline ML method is able to perform well for AIRR classification, we generated a collection of synthetic AIRR benchmark data sets encompassing a wide range of data set architecture-associated and immune state-associated sequence patterns (signal) complexity. We trained ≈1,700 ML models with varying assumptions regarding immune signal on ≈1,000 data sets with a total of ≈250,000 AIRRs containing ≈46 billion TCRβ CDR3 amino acid sequences, thereby surpassing the sample sizes of current state-of-the-art AIRR-ML setups by two orders of magnitude. We found that L1-penalized logistic regression achieved high prediction accuracy even when the immune signal occurs only in 1 out of 50,000 AIR sequences.

CONCLUSIONS

We provide a reference benchmark to guide new AIRR-ML classification methodology by (i) identifying those scenarios characterized by immune signal and data set complexity, where baseline methods already achieve high prediction accuracy, and (ii) facilitating realistic expectations of the performance of AIRR-ML models given training data set properties and assumptions. Our study serves as a template for defining specialized AIRR benchmark data sets for comprehensive benchmarking of AIRR-ML methods.

Fever as an evolutionary agent to select immune complexes interfaces

We herein analyzed all available protein–protein interfaces of the immune complexes from the Protein Data Bank whose antigens belong to pathogens or cancers that are modulated by fever in mammalian hosts. We also included, for comparison, protein interfaces from immune complexes that are not significantly modulated by the fever response. We highlight the distribution of amino acids at these viral, bacterial, protozoan and cancer epitopes, and at their corresponding paratopes that belong strictly to monoclonal antibodies. We identify the “hotspots”, i.e. residues that are highly connected at such interfaces, and assess the structural, kinetic and thermodynamic parameters responsible for complex formation. We argue for an evolutionary pressure for the types of residues at these protein interfaces that may explain the role of fever as a selective force for optimizing antibody binding to antigens.

Advances in computational structure-based antibody design

Antibodies are currently the most important class of biotherapeutics and are used to treat numerous diseases. Recent advances in computational methods are ushering in a new era of antibody design, driven in part by accurate structure prediction. Previously, structure-based antibody design has been limited to a relatively small number of cases where accurate structures or models of both the target antigen and antibody were available. As we move towards a time where it is possible to accurately model most antibodies and antigens, and to reliably predict their binding site, there is vast potential for true computational antibody design. In this review, we describe the latest methods that promise to launch a paradigm shift towards entirely in silico structure-based antibody design.

Next-Generation Molecular Discovery: From Bottom-Up In Vivo and In Vitro Approaches to In Silico Top-Down Approaches for Therapeutics Neogenesis

Protein and drug engineering comprises a major part of the medical and research industries, and yet approaches to discovering and understanding therapeutic molecular interactions in biological systems rely on trial and error. The general approach to molecular discovery involves screening large libraries of compounds, proteins, or antibodies, or in vivo antibody generation, which could be considered “bottom-up” approaches to therapeutic discovery. In these bottom-up approaches, a minimal amount is known about the therapeutics at the start of the process, but through meticulous and exhaustive laboratory work, the molecule is characterised in detail. In contrast, the advent of “big data” and access to extensive online databases and machine learning technologies offers promising new avenues to understanding molecular interactions. Artificial intelligence (AI) now has the potential to predict protein structure at an unprecedented accuracy using only the genetic sequence. This predictive approach to characterising molecular structure—when accompanied by high-quality experimental data for model training—has the capacity to invert the process of molecular discovery and characterisation. The process has potential to be transformed into a top-down approach, where new molecules can be designed directly based on the structure of a target and the desired function, rather than performing screening of large libraries of molecular variants. This paper will provide a brief evaluation of bottom-up approaches to discovering and characterising biological molecules and will discuss recent advances towards developing top-down approaches and the prospects of this.

An electronic biosensor based on semiconducting tetrazine polymer immobilizing matrix coated on rGO for carcinoembryonic antigen

Point-of-care devices are expected to play very critical roles in early diagnosis and better treatment of cancer. Here, we report the end-to-end development of novel and portable biosensors for detecting carcinoembryonic antigen (CEA), a cancer biomarker, almost instantly at room temperature. The device uses reduced graphene oxide (rGO) as the base conducting layer and a novel poly[(1,4-phenylene)-alt-(3,6-(1,2,4,5-tetrazine)/3,6-(1,2,4,5-dihydrotetrazine))] (PhPTz) as an immobilizing matrix for the CEA antibodies. Judiciously introduced nitrogen-rich semiconducting PhPTz brings multiple advantages to the device—(1) efficiently immobilizes anti-CEA via synergistic H-bonding with peptide and N-glycal units and (2) transports the charge density variations, originated upon antibody-antigen interactions, to the rGO layer. The CEA was dropped onto the anti-CEA/PhPTz/rGO devices at ambient conditions, to facilitate binding and the change in current flowing through the sensors was measured. A response of 2.75–33.7 μA was observed when the devices were tested for a broad range of concentrations (0.25 pg/mL to 800 ng/mL) of CEA. A portable read-out circuit was assembled using Arduino UNO and a voltage divider circuit, and a simple algorithm was developed for the classification of the CEA concentrations. The prediction accuracy of the interfacing electronics along with the algorithm was found to be 100%.

Antibody structure prediction using interpretable deep learning

Therapeutic antibodies make up a rapidly growing segment of the biologics market. However, rational design of antibodies is hindered by reliance on experimental methods for determining antibody structures. Here, we present DeepAb, a deep learning method for predicting accurate antibody F_V structures from sequence. We evaluate DeepAb on a set of structurally diverse, therapeutically relevant antibodies and find that our method consistently outperforms the leading alternatives. Previous deep learning methods have operated as “black boxes” and offered few insights into their predictions. By introducing a directly interpretable attention mechanism, we show our network attends to physically important residue pairs (e.g., proximal aromatics and key hydrogen bonding interactions). Finally, we present a novel mutant scoring metric derived from network confidence and show that for a particular antibody, all eight of the top-ranked mutations improve binding affinity. This model will be useful for a broad range of antibody prediction and design tasks.

•

DeepAb, a deep learning method for antibody structure, is presented

•

Structures from DeepAb are more accurate than alternatives

•

Outputs of DeepAb provide interpretable insights into structure predictions

•

DeepAb predictions should facilitate design of novel antibody therapeutics

Accurate structure models are critical for understanding the properties of potential therapeutic antibodies. Conventional methods for protein structure determination require significant investments of time and resources and may fail. Although greatly improved, methods for general protein structure prediction still cannot consistently provide the accuracy necessary to understand or design antibodies. We present a deep learning method for antibody structure prediction and demonstrate improvement over alternatives on diverse, therapeutically relevant benchmarks. In addition to its improved accuracy, our method reveals interpretable outputs about specific amino acids and residue interactions that should facilitate design of novel therapeutic antibodies.