Biology subjects

Yao, J.

Publications and source records attributed to Yao, J..

At least 109 records · Page 6Linked to original sources

bioRxiv2024

Lacticaseibacillus rhamnosus P118 enhances host tolerance to Salmonella infection by promoting microbe-derived indole metabolites

Salmonella is one of the most common foodborne pathogens, resulting in inflammatory gastroenteritis and frequently accompanied by dysbiosis. Gut commensals, such as Lactobacillus species, have been proven to exhibit broad anti-bacterial activities and protect hosts against pathogenic infections. Here, Lacticaseibacillus rhamnosus strain P118, screened from 290 isolates recovered from fermented yoghurts and healthy piglet intestines using traditional and Caenorhabditis elegans-infection screening strategies, exerts great probiotic properties. Notably, P118 and its supernatant exhibited great antibacterial activities and attenuated C. elegans susceptibility to Salmonella infection. We found that P118 protected mice against Salmonella lethal infections by enhancing colonization resistance, reducing pathogen invasion, alleviating intestinal pro-inflammatory response, and improving microbial dysbiosis and fecal metabolite changes. Microbiota and fecal metabolome analyses suggested P118 administration significantly decreased the relative abundances of potentially harmful microbes (e.g., Salmonella, Anaeroplasma, Klebsiella) and increased the fecal levels of tryptophan and its derivatives (indole, indole-3-acrylic acid, 5-hydroxytryptophan, 5-methoxyindoleacetate). Deterministic processes determined the gut microbial community assembly of P118-pretreated mice. Integrated omics further demonstrated that P118 probiotic activities in enhancing host tolerance to Salmonella infection were mediated by microbe-derived tryptophan/indole metabolites (e.g., indole-3-acrylic acid, indole, tryptophan, 5-methoxyindoleacetic acid, and 5-hydroxytryptophan). Collective results demonstrate that L. rhamnosus P118 could enhance host tolerance to Salmonella infections via various pathways, including direct antibacterial actions, inhibiting Salmonella colonization and invasion, attenuating pro-inflammatory responses of intestinal macrophages, and modulating gut microbiota mediated by microbe-derived indole metabolites.

microbiology↗

bioRxiv2024

An Artificial Intelligence Model for Profiling the Landscape of Antigen-binding Affinities of Massive BCR Sequencing Data

The interaction between antigens and antibodies (B cell receptors, BCRs) is the key step underlying the function of the humoral immune system in cancers and other biological contexts. The capability to profile the landscape of antigen-binding affinities of a vast number of BCRs will provide a powerful tool to reveal novel insights at unprecedented levels and will yield powerful tools for translational development. However, current experimental approaches for detecting antibody-antigen interactions are costly and time-consuming, and can only achieve low-to-mid throughput. On the other hand, bioinformatics tools in the field of antibody informatics mostly focus on optimization of antibodies given known binding antigens, which is a very different research question and of limited scope. In this work, we developed an innovative Artificial Intelligence tool, Cmai, to address the prediction of binding between antibodies and antigens that can be scaled to high-throughput sequencing data. We devised a biomarker metric based on the output from Cmai applied to high-throughput BCR sequencing data to predict the landscape of the antigen-binding affinities of the BCRs. We found that extracellular antigens on malignant tumor cells are inducing B cell infiltrations, and the infiltrating B cells have a greater tendency to co-localize with tumor cells expressing these antigens. We further found that the abundance of tumor antigen-targeting antibodies is predictive of immune-checkpoint inhibitor (ICI) treatment response. On the other hand, we found that, during immune-related adverse events (irAEs) caused by ICI treatment, humoral immunity is preferentially responsive to intracellular antigens from the organs affected by the irAEs. Overall, we developed a powerful tool for elucidating the antigen binding landscape of BCR repertoires, and it shall also propell antibody-centric translational applications for cancers and other diseases.

immunology↗

bioRxiv2024

CD-GPT: A Biological Foundation Model Bridging the Gap between Molecular Sequences Through Central Dogma

The central dogma serves as a fundamental framework for understanding the flow and expression of genetic information within living organisms, facilitating the connection of diverse biological sequences across molecule types. In this study, we present CD-GPT (Central Dogma Generative Pretrained Transformer), a generative biological foundation model with 1 billion parameters, aiming to capture the sequence relationships between DNA, RNA, and proteins. We model sequences in a unified representational space and employ a shared, multi-molecule vocabulary to narrow their distances in the embedding space effectively. Through extensive pretraining on nucleotide and amino acid sequence data, CD-GPT exhibits exceptional performance in a wide range of predictive and generative downstream tasks, including mono-molecular and multi-molecular analyses. Notably, CD-GPT excels in tasks such as genomic element detection, protein property prediction, RNA-protein interaction identification and also generative tasks like protein generation and reverse translation. The versatility of CD-GPT opens up promising avenues for advanced multi-omics analysis.

bioinformatics↗

bioRxiv2024

Human tau promotes Warburg effect-like glycolytic metabolism under acute hyperglycemia conditions through modulating the homeostasis of protein-membrane association

The neurofilaments formed by hyperphosphorylated tau is a hallmark of tau-related neurodegenerative disease, including Alzheimers disease, tau related FTDP-17, Picks disease, et al. However, the biological functions of tau and the physiological significance of its phosphorylation are still not fully understood. By using human tau (441 a.a.) transgenic (hTau) mice in which murine tau has been deleted simultaneously, murine tau knockout (Tau KO) mice and C57BL/6J (C57) mice, unexpectedly, we found that under acute hyperglycemia conditions, JNK but not previously reported GSK-3{beta} mediated tau phosphorylation. Moreover, Akt, the upstream GSK-3{beta} inhibitory kinase, was activated in a tau dependent manner. By comparing the membrane-associated proteome, we found that human tau influenced the homeostasis of protein-membrane association under acute hyperglycemia conditions. Of note, with respect to WT and Tau KO mice, the membrane-association of Krts, TFAM, TRAP1, mTOR et al, were strengthened by human tau. Whereas, the membrane-association of ribosomal proteins Rpls, proteasome proteins Psmds, and mitochondrial proteins, such as COXs, Ndufa1, Mtnt4, et al, were impeded by human tau. In vitro study showed that aerobic glycolysis was promoted in the presence of human tau, which maintained NAD+/NADH ratio. On the other hand, it restricted oxidative phosphorylation level, modulated the activity of SDH, and reduced ROS production upon challenging by high glucose. Furthermore, under acute high glucose conditions, the presence of human tau significantly augmented Akt activation, but inhibited 4EBP phosphorylation simultaneously, indicating that human tau is also involved in regulating the alternative activation of mTORC1/2. In summary, the current study revealed that human tau played an important role in regulating glycolytic metabolism under acute high hyperglycemia conditions, which is similar with the Warburg-effect, through influencing the homeostasis of protein-membrane association.

molecular biology↗

bioRxiv2024

Tumor microenvironment acidosis favors pancreatic cancer stem cell properties and in vivo metastasis

The acidic tumor microenvironment favors cancer aggressiveness via incompletely understood pathways. Here, we asked whether acidic environments select for cancer stem cell (CSC) properties. Bulk RNA-seq of Panc-1 human pancreatic cancer cells adapted to extracellular pH 6.5 revealed upregulation of CSC markers including CD44, EpCam, Nestin and aldehyde dehydrogenases, and CSC pathway enrichment. We therefore assessed CSC characteristics of acid-adapted (AA) and non-adapted (Ctrl) PaTu8988s and MiaPaca-2 pancreatic cancer cells. Compared to Ctrl, AA cells exhibited increased ALDH- and {beta}-catenin activity and pancreatosphere-forming efficiency, classical CSC characteristics. Panc-1, PaTu8988s and MiaPaCa-2 AA cells differed in CSC marker expression, and AA cells did not exhibit typical flow cytometric CSC populations. However, single-nucleus sequencing identified the acid adaptation-induced emergence of a population with clear CSC characteristics. Finally, in an orthotopic mouse model, AA Panc-1 cells drove strongly increased aggressiveness and liver metastasis compared to Ctrl cells. We conclude that acid-adaptation of pancreatic cancer cells leads to enrichment of a CSC phenotype with unusual traits, providing new insight into how acidic tumor microenvironments favor cancer aggressiveness.

cancer biology↗

bioRxiv2024

Incorporating Microbial Pilin-Based Nanowires into a Water-Stable Electronic Polymer Composite

Electrically conductive protein nanowires (e-PNs), microbially produced from a pilin monomer, are a novel, sustainable electronic material that can be genetically tailored for specific functions. e-PNs, expressed with Escherichia coli grown on the biodiesel byproduct glycerol, and mixed with polyvinyl butyral yielded a transparent, electrically conductive water-stable composite.Composite conductivity was adjusted by modifying the e-PN concentration or incorporating e-PNs genetically tuned for different conductivities. Electronic devices in which composites were the sensor component differentially responded to dissolved ammonia over a wide concentration range (1{micro}M-1M). Genetically modifying e-PNs to display an ammonia-binding peptide on their outer surface increased the sensor response to ammonia 10-fold. These results, coupled with the flexibility to design peptides for specific binding of diverse analytes, demonstrate that sustainably produced e-PNs offer the possibility of incorporating multiple sensor components, each specifically designed to detect different analytes with high sensitivity and selectivity, within one small sensor device.

bioengineering↗

bioRxiv2024

Genome sequencing of 'Fuji' apple clonal varieties reveals evolutionary history and genetic mechanism of the spur-type morphology

Somatic variations arising during asexual reproduction can lead to the emergence of bud sports exhibiting advantageous traits, forming the basis for bud sport breeding in perennial plants. Here, we report a near-complete, fully phased genome assembly of Fuji apple, enabling comprehensive identification of somatic variants across 74 clonally propagated Fuji varieties. Phylogenetic analysis of Fuji sport varieties indicates that the emergence of spur-type and early maturation traits results from multiple independent formation events. A number of putative functional somatic variants are identified, including one spur-type specific deletion located in the promoter of the TCP transcription factor gene MdTCP11. DNA methylation level of the deletion-associated miniature inverted-repeat transposable element is lower in the spur-type varieties compared to the standard-type varieties, while the expression of MdTCP11 is significantly higher. Overexpression of MdTCP11 in apple decreases the plant height, highlighting its important role in the development of spur-type apple varieties. This study sheds light on the 80-year cloning history of Fuji and provides valuable candidate causative mutations for apple breeding.

plant biology↗

bioRxiv2024

An integrated single-nucleus and spatial transcriptomics atlas reveals the molecular landscape of the human hippocampus

The hippocampus contains many unique cell types, which serve the structures specialized functions, including learning, memory and cognition. These cells have distinct spatial organization, morphology, physiology, and connectivity, highlighting the importance of transcriptome-wide profiling strategies that retain cytoarchitectural organization. Here, we generated spatially-resolved transcriptomics (SRT) and single-nucleus RNA-sequencing (snRNA-seq) data from adjacent tissue sections of the anterior human hippocampus in ten adult neurotypical donors to define molecular profiles for hippocampal cell types and spatial domains. Using non-negative matrix factorization (NMF) and label transfer, we integrated these data by defining gene expression patterns within the snRNA-seq data and inferring their expression in the SRT data. We identified NMF patterns that captured transcriptional variation across neuronal cell types and indicated that the response of excitatory and inhibitory postsynaptic specializations were prioritized in different SRT spatial domains. We used the NMF and label transfer approach to leverage existing rodent datasets, identifying patterns of activity-dependent transcription and subpopulations of dentate gyrus granule cells in our SRT dataset that may be predisposed to participate in learning and memory ensembles. Finally, we characterized the spatial organization of NMF patterns corresponding to non-cornu ammonis pyramidal neurons and identified snRNA-seq clusters mapping to distinct regions of the retrohippocampus, to three subiculum layers, and to a population of presubiculum neurons. To make this comprehensive molecular atlas accessible to the scientific community, both raw and processed data are freely available, including through interactive web applications.

neuroscience↗

bioRxiv2024

Highly conserved brain vascular receptor ALPL mediates transport of engineered viral vectors across the blood-brain barrier

Delivery of systemically administered therapeutics to the central nervous system (CNS) is restricted by the blood-brain barrier (BBB). Bioengineered Adeno-Associated Virus (AAV) capsids have been shown to penetrate the BBB with great efficacy in mouse and non-human primate models, but their translational potential is often limited by species selectivity and undefined mechanisms of action. Here, we apply our RNA-guided TRACER AAV capsid evolution platform to generate VCAP-102, an AAV9 variant with markedly increased brain tropism following intravenous delivery in both rodents and primates. VCAP-102 demonstrates a similar CNS tropism in cynomolgus macaque, african green monkey, marmoset and mouse, showing 20- to 400-fold increased transgene expression across multiple brain regions relative to AAV9. We demonstrate that the enhanced CNS tropism of VCAP-102 results from direct interaction with alkaline phosphatase (ALPL), a highly conserved membrane-associated protein expressed on the brain vasculature. VCAP-102 interacts with human, primate and murine ALPL isoforms, and ectopic expression of ALPL is sufficient to initiate receptor-mediated transcytosis of VCAP-102 in an in vitro transwell model. Our work identifies VCAP-102 as a cross-species CNS gene delivery vector with a strong potential for clinical translation and establishes ALPL as a brain delivery shuttle capable of efficient BBB transport to maximize CNS delivery of biotherapeutics.

neuroscience↗

bioRxiv2024

Translating Ethiopian potato seed networks: Identifying strategic intervention points for managing bacterial wilt and other diseases

ContextInformal seed trade can exacerbate crop disease epidemics. Potato production across Ethiopia is threatened by the spread of seedborne pests and pathogens, particularly by bacterial wilt, caused by the Ralstonia solanacearum Species Complex (RSSC). The RSSC is commonly dispersed through informal trade of seed potato, with the potential to move long distances across Ethiopia and among trading countries. Efficient disease testing programs and formal seed systems can reduce the risk of disease expansion in a countrys potato cropping system. ObjectiveIn this study, we characterized networks of potato value chain actors. We also identified candidate locations for disease surveillance and management interventions for improved seed systems, and locations at high risk for bacterial wilt establishment. We propose strategies to reduce the spread of bacterial wilt via infected seed. MethodsWe surveyed seed potato stakeholders to characterize interaction networks of potato value chain actors with a special focus on stakeholders engaged in seed potato quality assurance. We collected data regarding Ethiopias potato seed systems and analyzed the risk of spread of RSSC and other pathogens across the country through expert knowledge elicitation. Network metrics were used to characterize the informal seed trade system across Ethiopia, simulating pathogen spread across a network through scenario analyses. We characterized potato exports and imports to identify the risk of bacterial wilt movement through Ethiopias formal trading partners and neighboring countries where bacterial wilt has not yet been reported. ResultsWare potato farmers and traders were reported to have weak communication with other stakeholders in the potato value chain. In network analyses and simulated epidemics, locations in Agew Awi, Gamo, Gofa, Kembata and Tembaro zones were identified as candidate priorities for national surveillance of pathogen invasion and expansion through management interventions and formal seed system development. Ethiopia has formal trade with Sudan, Southern Sudan, Russia, and other countries where bacterial wilt has not been reported. Ethiopia may be at risk of reintroduction of the RSSC from countries where it is present, like Kenya and India. SignificanceImproving seed systems to manage R. solanacearum and other seedborne pathogens is important for supporting food security and the livelihoods of smallholder farmers in Ethiopia. Implementing surveillance systems and management programs in locations like those identified in Agew Awi, Gamo, Gofa, Kembata, and Tembaro zones, and improving the communication between ware potato traders and other stakeholders, can help to strengthen informal trade of seed potato and mitigate bacterial wilt spread in infected seed.

ecology↗

bioRxiv2024

Fast and accurate modeling and design of antibody-antigen complex using tFold

Accurate prediction of antibody-antigen complex structures holds significant potential for advancing biomedical research and the design of therapeutic antibodies. Currently, structure prediction for protein monomers has achieved considerable success, and promising progress has been made in extending this achievement to the prediction of protein complexes. However, despite these advancements, fast and accurate prediction of antibody-antigen complex structures remains a challenging and unresolved issue. Existing end-to-end prediction methods, which rely on homology and templates, exhibit sub-optimal accuracy due to the absence of co-evolutionary constraints. Meanwhile, conventional docking-based methods face difficulties in identifying the contact interface between the antigen and antibody and require known structures of individual components as inputs. In this study, we present a fully end-to-end approach for three-dimensional (3D) atomic-level structure predictions of antibodies and antibody-antigen complexes, referred to as tFold-Ab and tFold-Ag, respectively. tFold leverages a large protein language model to extract both intra-chain and inter-chain residue-residue contact information, as well as evolutionary relationships, avoiding the time-consuming multiple sequence alignment (MSA) search. Combined with specially designed modules such as the AI-driven flexible docking module, it achieves superior performance and significantly enhanced speed in predicting both antibody (1.6% RMSD reduction in the CDR-H3 region, thousand times faster) and antibody-antigen complex structures (37% increase in DockQ score, over 10 times faster), compared to AlphaFold-Multimer. Given the performance and speed advantages, we further extend the capability of tFold for structure-based virtual screening of binding antibodies, as well as de novo co-design of both structure and sequence for therapeutic antibodies. The experiment results demonstrate the potential of tFold as a high-throughput tool to enhance processes involved in these tasks. To facilitate public access, we release code and offer a web service for antibody and antigen-antibody complex structure prediction, which is available at https://drug.ai.tencent.com/en.

bioinformatics↗

bioRxiv2024

A single-cell 3D spatiotemporal multi-omics atlas from Drosophila embryogenesis to metamorphosis

The development of a multicellular organism is a highly intricate process tightly regulated by numerous genes and pathways in both spatial and temporal manners. Here, we present Flysta3D, a comprehensive multi-omics atlas of the model organism Drosophila, spanning its developmental lifespan from embryo to pupa. Our datasets encompass 3D single-cell spatial transcriptomic, single-cell transcriptomic, and single-cell chromatin accessibility information. By integrating these multi-dimensional data, we constructed cell state trajectories that uncover the detailed profiles of tissue development. With a focus on the central nervous system (CNS) and midgut, we dissected the spatiotemporal dynamics of gene regulatory networks, cell type diversity, and morphological changes from a multi-omics perspective. This extensive atlas provides an unprecedentedly rich resource and serves as a systematic platform for studying Drosophila development with integrated single-cell data at an ultra-high spatiotemporal resolution.

developmental biology↗

bioRxiv2024

Spatial domain detection using contrastive self-supervised learning for spatial multi-omics technologies

Recent advances in spatially-resolved single-omics and multi-omics technologies have led to the emergence of computational tools to detect or predict spatial domains. Additionally, histological images and immunofluorescence (IF) staining of proteins and cell types provide multiple perspectives and a more complete understanding of tissue architecture. Here, we introduce Proust, a scalable tool to predict discrete domains using spatial multi-omics data by combining the low-dimensional representation of biological profiles based on graph-based contrastive self-supervised learning. Our scalable method integrates multiple data modalities, such as RNA, protein, and H&E images, and predicts spatial domains within tissue samples. Through the integration of multiple modalities, Proust consistently demonstrates enhanced accuracy in detecting spatial domains, as evidenced across various benchmark datasets and technological platforms.

bioinformatics↗

bioRxiv2023

Perception of butenolides by Bacillus subtilis via the α/β-hydrolase RsbQ

The regulation of behavioural and developmental decisions by small molecules is common to all domains of life. In plants, strigolactones and karrikins are butenolide growth regulators that influence several aspects of plant growth and development, as well as interactions with symbiotic fungi1-3. DWARF14 (D14) and KARRIKIN INSENSITIVE2 (KAI2) are homologous receptors that perceive strigolactones and karrikins, respectively, and that hydrolyse their ligands to effect signal transduction4-7. RsbQ, a homologue of D14 and KAI2 from the Gram-positive bacterium Bacillus subtilis, regulates growth responses to nutritional stress via the alternative transcription factor SigmaB ({sigma}B)8,9. However, the molecular function of RsbQ is unknown. Here we show that RsbQ perceives butenolide compounds that are bioactive in plants. RsbQ is thermally destabilised by the synthetic strigolactone GR24 and its desmethyl butenolide equivalent dGR24. We show that, like D14 and KAI2, RsbQ is a functional butenolide hydrolase that undergoes covalent modification of the catalytic histidine residue. Exogenous application of both GR24 and dGR24 inhibited the endogenous signalling function of RsbQ in vivo, with dGR24 being 10-fold more potent. Application of dGR24 to B. subtilis phenocopied loss-of-function rsbQ mutations and led to a significant down-regulation of {sigma}B-regulated transcripts. We also discovered that exogenous butenolides promoted the transition from planktonic to biofilm growth. Our results suggest that butenolides may serve as inter-kingdom signalling compounds between plants and bacteria to help shape rhizosphere communities.

microbiology↗

bioRxiv2023

Shared architectural patterns across the human cortical mantle predict visual representations and capture behavior across the lifespan

The microstructure of cells within human cerebral cortex varies across the cortical ribbon, where changes in cytoarchitecture and myeloarchitecture are thought to endow each region of cortex with its unique function. While fine-scale relative to a cell, these changes at population level impact architectural properties of cortex measurable in vivo by noninvasive MRI, such as the thickness and myelin content of cortex. This raises the question of whether or not we can use these in vivo architectural measures to understand cortical organization, function, and development more broadly. Using human visual cortex as a test bed, we demonstrated two architectural gradients, one in which cytoarchitecture and myeloarchitecture converge and another in which they diverge. These two gradients underlie the structural and functional topography of visual cortex, even predicting the presence of new visual representations. Moreover, the two gradients show distinct visual behavior relevance and lifespan trajectory. These findings provide a more general framework for understanding human cortex, showing that architectural gradients are a measurable fingerprint of functional organization and ontogenetic routines in the human brain.

neuroscience↗

bioRxiv2023

TCM-CMap: A Novel Computational Methodology to Link Gene Perturbation to Therapeutic Targets in Complex TCM Prescriptions

Herbal medicine, especially Traditional Chinese medicine (TCM), is a valuable resource of natural products for drug discovery with obvious therapeutic effects. However, the mechanisms of action (MOAs) of herbal medicine are often unknown due to limited target information and the complexity of multiple ingredients and targets. This study developed a Herb-CMap algorithm to prioritize active ingredients and targets within herbal medicine by integrating transcriptomics- based gene perturbation data with a random walk algorithm. This methodology bridges the gap between gene perturbation of herbal medicine and its therapeutic target. Using the Suhuang antitussive capsule (Suhuang) for treating cough variant asthma (CVA) as a case study, we identify and experimentally verify that quercetin and luteolin directly interact with Il17a, Pik3cb, Pik3cd, Akt1, and Tnf. These interactions inhibit the IL-17 signaling pathway and inactivate PI3K, Akt, and NF- {kappa}B, preventing lung inflammation and treating CVA. Our findings demonstrate the potential of the Herb-CMap methodology to provide insights into the molecular MOAs of herbal medicine, thereby advancing drug discovery from herbal medicine.

systems biology↗

bioRxiv2023

De novo generation of T-cell receptors with desired epitope-binding property by leveraging a pre-trained large language model

Generating T-cell receptors (TCRs) with desired epitope-binding properties is a fundamental step in the development of immunotherapies, yet heavily relies on laborious and expensive wet experiments. Recent advancements in generative artificial intelligence have demonstrated promising power in protein design and engineering. In this regard, we propose a large language model, termed Epitope-Receptor-Transformer (ERTransformer), for the de novo generation of TCRs with the desired epitope-binding property. ERTransformer is built on EpitopeBERT and ReceptorBERT, which are trained using 1.9 million epitope sequences and 33.1 million TCR sequences, respectively. To demonstrate the model capability, we generate 1000 TCRs for each of the five epitopes with known natural TCRs. The artificial TCRs exhibit low sequence identity (average Bit-score 27.64 with a standard deviation of 1.50) but high biological function similarity (average BLOSUM62 score 32.32 with a standard deviation of 12.01) to natural TCRs. Furthermore, the artificial TCRs are not very structurally identical to natural ones (average RMSD 2.84 [A] with a standard deviation of 1.21 [A]) but exhibit a comparable binding affinity towards the corresponding epitopes. Our work highlights the tremendous potential of applying ERTransformer to generate novel TCRs with desired epitope-binding ability.

bioinformatics↗

bioRxiv2023

De novo generation of antibody CDRH3 with a pre-trained generative large language model

Artificial Intelligence (AI) techniques have made great advances in assisting antibody design. However, antibody design still heavily relies on isolating antigen-specific antibodies from serum, which is a resource-intensive and time-consuming process. To address this issue, we propose a Pre-trained Antibody generative large Language Model (PALM) for the de novo generation of artificial antibodies heavy chain complementarity-determining region 3 (CDRH3) with desired antigen-binding specificity, reducing the reliance on natural antibodies. We also build a high-precision model antigen-antibody binder (A2binder) that pairs antigen epitope sequences with antibody sequences to predict binding specificity and affinity. PALM-generated antibodies exhibit binding ability to SARS-CoV-2 antigens, including the emerging XBB variant, as confirmed through in-silico analysis and in-vitro assays. The in-vitro assays validated that PALM-generated antibodies achieve high binding affinity and potent neutralization capability against both wild-type and XBB spike proteins of SARS-CoV-2. Meanwhile, A2binder demonstrated exceptional predictive performance on binding specificity for various epitopes and variants. Furthermore, by incorporating the attention mechanism into the PALM model, we have improved its interpretability, providing crucial insights into the fundamental principles of antibody design.

bioinformatics↗