DA-TransUNet: integrating spatial and channel dual attention with transformer U-net for medical image segmentation.

Accurate medical image segmentation is critical for disease quantification and treatment evaluation. While traditional U-Net architectures and their transformer-integrated variants excel in automated segmentation tasks. Existing models also struggle with parameter efficiency and computational complexity, often due to the extensive use of Transformers. However, they lack the ability to harness the image's intrinsic position and channel features. Research employing Dual Attention mechanisms of position and channel have not been specifically optimized for the high-detail demands of medical images. To address these issues, this study proposes a novel deep medical image segmentation framework, called DA-TransUNet, aiming to integrate the Transformer and dual attention block (DA-Block) into the traditional U-shaped architecture. Also, DA-TransUNet tailored for the high-detail requirements of medical images, optimizes the intermittent channels of Dual Attention (DA) and employs DA in each skip-connection to effectively filter out irrelevant information. This integration significantly enhances the model's capability to extract features, thereby improving the performance of medical image segmentation. DA-TransUNet is validated in medical image segmentation tasks, consistently outperforming state-of-the-art techniques across 5 datasets. In summary, DA-TransUNet has made significant strides in medical image segmentation, offering new insights into existing techniques. It strengthens model performance from the perspective of image features, thereby advancing the development of high-precision automated medical image diagnosis. The codes and parameters of our model will be publicly available at https://github.com/SUN-1024/DA-TransUnet.

Deciphering the antigen specificities of antibodies by clustering their complementarity determining region sequences.

IMPORTANCE: Determining antigen and epitope specificity is an essential step in the discovery of therapeutic antibodies as well as in the analysis adaptive immune responses to disease or vaccination. Despite extensive efforts, deciphering antigen specificity solely from BCR amino acid sequence remains a challenging task, requiring a combination of experimental and computational approaches. Here, we describe and experimentally validate a simple and straightforward approach for grouping antibodies that share antigen and epitope specificities based on their CDR sequence similarity. This approach allows us to identify the specificities of a large number of antibodies whose antigen targets are unknown, using a small fraction of antibodies with well-annotated binding specificities.

Structural Modeling of Adaptive Immune Responses to Infection.

Antibody and TCR modeling are becoming important as more and more sequence data becomes available to the public. One of the pressing questions now is how to use such data to understand adaptive immune responses to disease. Infectious disease is of particular interest because the antigens driving such responses are often known to some extent. Here, we describe tips for gathering data and cleaning it for use in downstream analysis. We present a method for high-throughput structural modeling of antibodies or TCRs using Repertoire Builder and its extensions. AbAdapt is an extension of Repertoire Builder for antibody-antigen docking from antibody and antigen sequences. ImmuneScape is a corresponding extension for TCR-pMHC 3D modeling. Together, these pipelines can help researchers to understand immune responses to infection from a structural point of view.

The evolution of structural genomics.

Structural genomics began as a global effort in the 1990s to determine the tertiary structures of all protein families as a response to large-scale genome sequencing projects. The immediate outcome was an influx of tens of thousands of protein structures, many of which had unknown functions. At the time, the value of structural genomics was controversial. However, the structures themselves were only the most obvious output. In addition, these newly solved structures motivated the emergence of huge data science and infrastructure efforts, which, together with advances in Deep Learning, have brought about a revolution in computational molecular biology. Here, we review some of the computational research carried out at the Protein Data Bank Japan (PDBj) during the Protein 3000 project under the leadership of Haruki Nakamura, much of which continues to flourish today.

Advances in antibody discovery from human BCR repertoires.

Antibodies make up an important and growing class of compounds used for the diagnosis or treatment of disease. While traditional antibody discovery utilized immunization of animals to generate lead compounds, technological innovations have made it possible to search for antibodies targeting a given antigen within the repertoires of B cells in humans. Here we group these innovations into four broad categories: cell sorting allows the collection of cells enriched in specificity to one or more antigens; BCR sequencing can be performed on bulk mRNA, genomic DNA or on paired (heavy-light) mRNA; BCR repertoire analysis generally involves clustering BCRs into specificity groups or more in-depth modeling of antibody-antigen interactions, such as antibody-specific epitope predictions; validation of antibody-antigen interactions requires expression of antibodies, followed by antigen binding assays or epitope mapping. Together with innovations in Deep learning these technologies will contribute to the future discovery of diagnostic and therapeutic antibodies directly from humans.

Landscape of infection enhancing antibodies in COVID-19 and healthy donors.

To assess the frequency of SARS-CoV-2 infection in the general population, we searched over 64 million heavy chain antibody sequences from healthy unvaccinated, healthy BNT162b2 vaccinated and COVID-19 patient repertoires for sequences similar to 11 previously reported enhancing antibodies. Although the distribution of sequence identities was similar in all three groups of repertoires, the COVID-19 and healthy vaccinated hits were significantly more clonally expanded than healthy unvaccinated hits. Furthermore, among the tested hits, 17 out of 94 from COVID-19 and 9 out of 59 from healthy vaccinated, compared with only 2 out of 96 from healthy unvaccinated, bound to the enhancing epitope. A total of 9 of the 28 epitope-binding antibodies enhanced ACE2 receptor binding to the spike protein. Together, this study revealed that infection enhancing-like antibodies are far more frequent in COVID-19 patients or healthy vaccinated donors than in healthy unvaccinated donors, but a reservoir of potential enhancing antibodies exists in healthy donors that could potentially mature to actual enhancing antibodies upon infection.

Improved Antibody-Specific Epitope Prediction Using AlphaFold and AbAdapt.

Antibodies recognize their cognate antigens with high affinity and specificity, but the prediction of binding sites on the antigen (epitope) corresponding to a specific antibody remains a challenging problem. To address this problem, we developed AbAdapt, a pipeline that integrates antibody and antigen structural modeling with rigid docking in order to derive antibody-antigen specific features for epitope prediction. In this study, we systematically assessed the impact of integrating the state-of-the-art protein modeling method AlphaFold with the AbAdapt pipeline. By incorporating more accurate antibody models, we observed improvement in docking, paratope prediction, and prediction of antibody-specific epitopes. We further applied AbAdapt-AF in an anti-receptor binding domain (RBD) antibody complex benchmark and found AbAdapt-AF outperformed three alternative docking methods. Also, AbAdapt-AF demonstrated higher epitope prediction accuracy than other tested epitope prediction tools in the anti-RBD antibody complex benchmark. We anticipate that AbAdapt-AF will facilitate prediction of antigen-antibody interactions in a wide range of applications.

AbAdapt: an adaptive approach to predicting antibody-antigen complex structures from sequence.

Motivation: The scoring of antibody-antigen docked poses starting from unbound homology models has not been systematically optimized for a large and diverse set of input sequences. Results: To address this need, we have developed AbAdapt, a webserver that accepts antibody and antigen sequences, models their 3D structures, predicts epitope and paratope, and then docks the modeled structures using two established docking engines (Piper and Hex). Each of the key steps has been optimized by developing and training new machine-learning models. The sequences from a diverse set of 622 antibody-antigen pairs with known structure were used as inputs for leave-one-out cross-validation. The final set of cluster representatives included at least one 'Adequate' pose for 550/622 (88.4%) of the queries. The median (interquartile range) ranks of these 'Adequate' poses were 22 (5-77). Similar results were obtained on a holdout set of 100 unrelated antibody-antigen pairs. When epitopes were repredicted using docking-derived features for specific antibodies, the median ROC AUC increased from 0.679 to 0.720 in cross-validation and from 0.694 to 0.730 in the holdout set. Availability and implementation: AbAdapt and related data are available at https://sysimm.org/abadapt/. Supplementary information: Supplementary data are available at Bioinformatics Advances online.

Flexible, Functional, and Familiar: Characteristics of SARS-CoV-2 Spike Protein Evolution.

The SARS-CoV-2 S protein is a major point of interaction between the virus and the human immune system. As a consequence, the S protein is not a static target but undergoes rapid molecular evolution. In order to more fully understand the selection pressure during evolution, we examined residue positions in the S protein that vary greatly across closely related viruses but are conserved in the subset of viruses that infect humans. These "evolutionarily important" residues were not distributed evenly across the S protein but were concentrated in two domains: the N-terminal domain and the receptor-binding domain, both of which play a role in host cell binding in a number of related viruses. In addition to being localized in these two domains, evolutionary importance correlated with structural flexibility and inversely correlated with distance from known or predicted host receptor-binding residues. Finally, we observed a bias in the composition of the amino acids that make up such residues toward more human-like, rather than virus-like, sequence motifs.

Methods for sequence and structural analysis of B and T cell receptor repertoires.

B cell receptors (BCRs) and T cell receptors (TCRs) make up an essential network of defense molecules that, collectively, can distinguish self from non-self and facilitate destruction of antigen-bearing cells such as pathogens or tumors. The analysis of BCR and TCR repertoires plays an important role in both basic immunology as well as in biotechnology. Because the repertoires are highly diverse, specialized software methods are needed to extract meaningful information from BCR and TCR sequence data. Here, we review recent developments in bioinformatics tools for analysis of BCR and TCR repertoires, with an emphasis on those that incorporate structural features. After describing the recent sequencing technologies for immune receptor repertoires, we survey structural modeling methods for BCR and TCRs, along with methods for clustering such models. We review downstream analyses, including BCR and TCR epitope prediction, antibody-antigen docking and TCR-peptide-MHC Modeling. We also briefly discuss molecular dynamics in this context.

Reversible control of enantioselectivity by the length of ketone substituent in biocatalytic reduction.

Enzyme engineering has been widely employed to tailor the substrate specificity and enantioselectivity of enzymes. In this study, we mutated Trp288, an unconserved residue in the small binding pocket of an acetophenone reductase from Geotrichum candidum NBRC 4597 (GcAPRD). Trp288 mutants showed substrate specificity expansion towards bulky-bulky ketones and enantioselectivity alteration which was highly dependent on the substrate substituent length. In aliphatic ketone reduction, enantioselectivity inverted from (S) to (R) when one of the substituents to the carbonyl carbon was elongated from propyl to butyl or pentyl. The best (R)-selective mutant, Trp288Val, achieved the reduction of 3-heptanone to its corresponding (R)-alcohol with 97% ee. Our docking simulation suggested that when enantioselectivity inverted to (R), only pro-R binding poses were productive. Gly94 played an important role to stabilize the butyl or pentyl group for their productive pro-R poses. Interestingly, when the substituent was further elongated, the enantioselectivity inverted back to the (S) form.

Correction to: Structural basis for a highly (S)-enantioselective reductase towards aliphatic ketones with only one carbon difference between side chain.

The original version of this article contains error for some of the authors corrections were not included during correction stage.

Structural basis for a highly (S)-enantioselective reductase towards aliphatic ketones with only one carbon difference between side chain.

Aliphatic ketones, such as 2-butanone and 3-hexanone, with only one carbon difference among side chains adjacent to the carbonyl carbon are difficult to be reduced enantioselectively. In this study, we utilized an acetophenone reductase from Geotrichum candidum NBRC 4597 (GcAPRD) to reduce challenging aliphatic ketones such as 2-butanone (methyl ethyl ketone) and 3-hexanone (ethyl propyl ketone) to their corresponding (S)-alcohols with 94% ee and > 99% ee, respectively. Through crystallographic structure determination, it was suggested that residue Trp288 limit the size of the small binding pocket. Docking simulations imply that Trp288 plays an important role to form a C-H⋯π interaction for proper orientation of ketones in the pro-S binding pose in order to produce (S)-alcohols. The excellent (S)-enantioselectivity is due to a non-productive pro-R binding pose, consistent with the observation that the (R)-alcohol acts as an inhibitor of (S)-alcohol oxidation.

Structural Modeling of Lymphocyte Receptors and Their Antigens.

Structural modeling plays a key role in protein function prediction on a genome-wide scale. For B and T lymphocyte receptors, the critical functional question is: which antigens and epitopes are targeted? With emerging B cell receptor (BCR) and T cell receptor (TCR) sequencing methods improving in both breadth and depth, there is a growing need for methods that can help answer this question. Since lymphocyte-antigen recognition depends on complementarity, structural modeling is likely to play an important role in understanding antigen specificity and affinity. In the case of BCRs, such modeling methods have a long history in the study and design of antibodies. However, for TCRs there are relatively few publicly available modeling tools, and, to our knowledge, none that incorporate interaction between TCRs and peptide-MHC (pMHC) complexes. Here, we provide a web-based tool, ImmuneScape ( https://sysimm.org/immune-scape/ ), to carry out TCR-pMHC modeling as a first step toward structure-based function prediction.