Biology subjects

Chen, Y.-Q.

Publications and source records attributed to Chen, Y.-Q..

8 recordsLinked to original sources

bioRxiv2026

AMPBAN: A Deep Learning Framework Integrating Protein Sequence and Structural Features for Antimicrobial Peptide Prediction

The escalating crisis of antimicrobial resistance poses a devastating and immediate threat to human life. Antimicrobial peptides (AMPs) are a promising antibiotic substitute to combat antimicrobial resistance. Compared with the traditional wet-lab screening approaches, computational models have largely improved the efficiency of predicting antimicrobial peptide. However, most computational models overlook or underutilize the evolution and structural information of peptides, which is crucial for understanding the peptide functions. Here, we proposed a sophisticated deep learning model to predict AMPs, Antimicrobial Peptide Bilinear Attention Network (AMPBAN), which incorporates peptide evolution features from ESM3 protein language model, structure features from ESMFold predicted with equivariant graph neural network (EGNN), and the joint information from sequence and structure learned via Bilinear Attention Network. AMPBAN consistently demonstrated superior accuracy and generalization compared to nine state-of-the-art AMP prediction models across multiple independent benchmarks. Furthermore, an ablation study confirms that our multimodal fusion strategy significantly refines the integration of sequence and structural signals, yielding superior predictive balance over single-modality models. This framework provides a robust tool for the accelerated discovery of novel AMPs and the advancement of next-generation antimicrobial drug development. The datasets, source code and models are available at https://github.com/baiwenhuim/ampban.

bioinformatics↗

bioRxiv2025

A Novel 'Three-in-one' Mucosal Vaccine Elicits Broad Protection against Three Distinct Clusters of ACE2-using Sarbecoviruseses

With the persistent risk of coronavirus epidemics, the development of universal vaccines is urgently needed and challenging to achieve. In this study, we aimed to design a multivalent immunogen to provide broad protection against diverse ACE2-using sarbecoviruses, which pose a significant threat to global health. Our analysis revealed that the receptor-binding domains (RBDs) of ACE2-using sarbecoviruses segregate into three antigenically distinct clusters, including one cluster that encompasses viruses related to SARS-CoV-2. Based on these findings, we engineered a three-in-one immunogen, designated as 3Rs-NC, which incorporates representative RBDs from all three clusters. 3Rs-NC preserved the natural epitope configuration of each monomeric RBD component, and efficiently elicited high levels of neutralizing antibodies against representative viruses and closely related sarbecoviruses. When administered intranasally with flagellin-based mucosal adjuvant KFD, 3Rs-NC induced robust and durable RBD-specific serum IgG and mucosal IgA responses in mice. Furthermore, KFD-adjuvanted 3Rs-NC conferred sustained protection in both the upper and lower respiratory tracts against SARS-CoV-2 Omicron BA.1 and SARS-like coronavirus WIV1. Additionally, 3Rs-NC immunization protected mice from lethal challenge of SARS-like coronavirus rRsSHC014S, with more efficient protection observed in female mice than male mice. This needle-free three-in-one vaccine represents a promising candidate for a universal mucosal vaccine against ACE2-using sarbecoviruses and could serve as a foundational component of three-in-one vaccine series to form a comprehensive coronavirus vaccine package. HighlightO_LInovel three-in-one immunogen 3Rs-NC induced a high level of neutralizing antibodies against three distinct ACE2-using Sabecovirus clusters. C_LIO_LIadministration of 3Rs-NC plus KFD adjuvant (3Rs-NC+KFDi.n) generated potent and long-lasting RBD-specific serum IgG and mucosal IgA. C_LIO_LI3Rs-NC+KFDi.n provided long-term protection in both upper- and lower-respiratory tracts in mice. C_LIO_LI3Rs-NC+KFDi.n could serve as a foundational component for next-generation coronavirus vaccine strategies. C_LI

immunology↗

bioRxiv2025

Predicting the Evolutionary and Functional Landscapes of Viruses with a Unified Nucleotide-Protein Language Model: LucaVirus

Predicting viral evolution and function remains a central challenge in biology, hindered by high sequence divergence and limited knowledge compared to cellular organisms. Here, we introduce LucaVirus, a multi-modal foundation model for viruses, trained on 25.4 billion nucleotide and amino acid tokens covering nearly all known viruses. LucaVirus learns biologically meaningful representations capturing relationships between sequences, protein/gene homology, and evolutionary divergence. Using these embeddings, we developed downstream models that address key virology tasks: identifying hidden viruses in genomic "dark matter", annotating enzymatic activities of uncharacterized proteins, predicting viral evolvability, and identifying antibody candidates for emerging viruses. LucaVirus achieves state-of-the-art results in three tasks and matches leading models in the fourth with one-third the parameters. Together, these findings demonstrate the power of a unified foundation model to comprehensively decode the viral world and establish LucaVirus as an efficient and versatile platform for AI-driven virology, from virus discovery to functional and therapeutic predictions.

bioinformatics↗

bioRxiv2024

Deep generative model for protein subcellular localization prediction

Protein sequence determines not only its structure but also its subcellular localization. Although a series of artificial intelligence models have been reported to predict protein subcellular localization, most of them provide only textual outputs. Here, we present deepGPS, a deep generative model for protein subcellular localization prediction. After trained with both protein primary sequences and protein subcellular localization fluorescence images, deepGPS shows the ability to predict cytoplasmic and nuclear localizations by reporting both textual labels and generative images as outputs. In addition, deepGPS shows potential to be further extended for other types of subcellular localization prediction, even with limited input data volumes for training. Finally, an openGPS website (https://bits.fudan.edu.cn/opengps) is constructed to provide a public and convenient platform for protein subcellular localization prediction with the scientific community.

bioinformatics↗

bioRxiv2023

An Empirical Study on the Motivation of Helping Behavior in Rats

Using rodent models to explore the motivation of helping behaviors has become a new trend in recent years. Empathy, the alleviation of personal distress, and desire for social contact have been considered motivations for rodents to engage in helping behaviors. We used 108 Sprague-Dawley rats as subjects and modified the two-chambered helping behavior experimental setup in Carvalheiro and colleagues study to explore the main motivations of helping behavior in rodents through three experiments. The findings suggest that (1) the desire for social contact and pursuit of an interesting environment are the primary motivations for helping behavior, regardless of the presence of a dark chamber, and (2) the alleviation of personal distress and prior experience of social contact rather than distress experience contribute to the onset and persistence of helping behavior.

animal behavior and cognition↗

bioRxiv2023

Spike substitution T813S increases Sarbecovirus fusogenicity by enhancing the usage of TMPRSS2.

SARS-CoV Spike (S) protein shares considerable homology with SARS-CoV-2 S, especially in the conserved S2 subunit (S2). S protein mediates coronavirus receptor binding and membrane fusion, and the latter activity can greatly influence coronavirus infection. We observed that SARS-CoV S is less effective in inducing membrane fusion compared with SARS-CoV-2 S. We identify that S813T mutation is sufficient in S2 interfering with the cleavage of SARS-CoV-2 S by TMPRSS2, reducing spike fusogenicity and pseudoparticle entry. Conversely, the mutation of T813S in SARS-CoV S increased fusion ability and viral replication. Our data suggested that residue 813 in the S was critical for the proteolytic activation, and the change from threonine to Serine at 813 position might be an evolutionary feature adopted by SARS-2-related viruses. This finding deepened the understanding of Spike fusogenicity and could provide a new perspective for exploring Sarbecovirus evolution. Author SummaryThe Spike strain of SARS-CoV-2 has accumulated many mutations during its time in circulation, most of which have occurred in the S1 region, and more specifically in the RBD, in an effort to either improve the viruss affinity for the receptor ACE2 or to enhance its ability to evade the immune system. Mutations in the Spike S2 region have more far-reaching effects than those in the S1 region because it is more conserved across sarbecoviruses. By comparing SARS and SARS2, we found that an important substitution at amino acid position 813 in the S2 region (T813S) disrupts the utilization of TMPRSS2 and can significantly influence viral entry into cells. This discovery deepens our knowledge of S proteins and provides new prospects for tracing the evolution of Sarbecoviruses.

microbiology↗

bioRxiv2021

A Safe and Effective Mucosal RSV Vaccine Consisting of RSV Phosphoprotein and Flagellin Variant

Respiratory syncytial virus (RSV) is a major cause of serious acute lower respiratory tract infection in infants and the elderly. No licensed RSV vaccine available thus far calls for the development of vaccines with new target(s) and vaccination strategies. Here, we constructed a recombinant protein, designated P-KFD1, comprised of RSV phosphoprotein (P) and E. coli K12 strain-derived flagellin variant KFD1. Intranasal (i.n.) immunization with P-KFD1 inhibits RSV replication in both upper and lower respiratory tract, and protects mice against lung disease without vaccine-enhanced disease (VED). The P-specific CD4+ T cells provoked by P-KFD1 i.n. immunization, either reside in or migrate to respiratory tract, mediate protection against RSV infection. Sc-RNA seq and carboxyfluorescein succinimidyl ester (CFSE) labeled cell transfer further characterized the Th1 and Th17 responses induced by P-KFD1. Finally, we found the anti-viral protection depends on either IFN-{gamma} or IL-17A. Collectively, P-KFD1 is promising as a safe and effective mucosal vaccine candidate to prevent RSV infection. HIGHLIGHTSO_LIA new subunit RSV vaccine candidate with new target and vaccination strategy, P-KFD1, is designed and generated C_LIO_LIIntranasal immunization with P-KFD1 protects mice against RSV infection and averts vaccine-enhanced disease C_LIO_LISc-RNA seq and CFSE-labelled cell transfer identified characteristics of the protective CD4+ T cells C_LIO_LILocal and peripheral CD4+ T cells provide protection against RSV infection dependent on either IFN-{gamma} or IL-17A C_LI

immunology↗

bioRxiv2018

Hemagglutinin stalk-reactive antibodies interfere with influenza virus neuraminidase activity by steric hindrance

AbstractHemagglutinin (HA) stalk-reactive antibodies are the basis of several current \"one-shot\" universal influenza vaccine efforts because they protect against a wide spectrum of influenza virus strains. The appreciated mechanism of protection by HA-stalk antibodies is to inhibit HA stalk reconfiguration, blocking viral fusion and entry. This study shows that HA stalk-reactive antibodies also inhibit neuraminidase (NA) enzymatic activity, prohibiting viral egress. NA inhibition (NI) is evident for an attached substrate but not for unattached small molecule cleavage of sialic acid. This suggests that the antibodies inhibit NA enzymatic activity through steric hindrance, thus limiting NA access to sialic acids when adjacent to HA on whole virions. Consistently, F(ab)2 fragments that occupy reduced area without loss of avidity or disrupted HA/NA interactions show significantly reduced NI activity. Notably, HA stalk binding antibodies lacking NI activity were unable to neutralize viral infection via microneutralization assays. This work suggests that NI activity is an important component of HA-stalk antibody mediated protection.\n\nSummaryThis study reports a new mechanism of protection that is mediated by influenza hemagglutinin-stalk reactive antibodies: inhibition of neuraminidase activity by steric hindrance, blocking access of neuraminidase to sialic acids when it is abutted next to hemagglutinin on whole virions.

microbiology↗