Biology subjects

Hou, T.

Publications and source records attributed to Hou, T..

At least 19 recordsLinked to original sources

bioRxiv2026

LucaCell: a sequence-centric foundation model for cross-species single-cell analysis

Single-cell foundation models have transformed transcriptomic analysis, yet most rely on fixed gene identifiers that limit transfer across species and data types. Here we present LucaCell, a sequence-centric foundation model that represents genes through pre-trained mRNA sequence embeddings rather than static gene annotations. Gene expression is discretized into bins and modeled with a Transformer encoder, enabling sequence-informed cell representation without a fixed gene-ID vocabulary. Pre-training on 85 million human and mouse single cells, LucaCell is evaluated on human, mouse and lemur gene expression profiles, human chromatin accessibility data, unaligned reads from more than 50 prokaryotic taxa, and five influenza A virus genomes. LucaCell enables manual-mapping-free cross-species cell type annotation and an alignment-free microbial embedding framework that simultaneously distinguishes bacterial species identity and intra-species physiological states. It also improves gene expression reconstruction by incorporating donor-specific exonic SNP information into mRNA sequence embeddings, and predicts cellular viral load across influenza A virus strains while highlighting infection-like transcriptional states in mock-infected cells. These results show that sequence-informed gene representation can improve the generalization of single-cell foundation models across species, data types, and predictive tasks.

bioinformatics↗

bioRxiv2026

BBBP_Atlas: Unified Interpretable Modeling of Blood Brain Barrier Permeability across Small Molecules and Peptides

Accurate prediction of blood-brain barrier permeability (BBBP) is essential for central nervous system drug discovery, yet existing models are often limited by their reliance on predefined physicochemical descriptors, small-molecule-centered training sets, or conformation-dependent representations, which restricts their transferability across chemically diverse modalities especially peptides. In addition, publicly available BBBP datasets remain fragmented, inconsistently standardized, and weakly controlled for molecular redundancy, increasing the risk of data leakage and overestimated model performance. In this study, we propose BBBP-Atlas, a structure-aware BBB permeability prediction model designed for unified modeling of small molecules and peptides with the first cross-modal dataset OmniBBBP. Designed to bypass descriptor and conformation dependencies, our model represents standardized molecular structures as atom-level graphs to capture local atom-bond environments and long-range topological dependencies associated with BBB transport. This design enables direct learning of structure-permeability relationships from molecular topology. For model training and evaluation, we curated a cross-modal, redundancy-filtered database OmniBBBP that seamlessly unifies small molecules and complex peptides, containing 10,218 unique compounds with 9,316 small molecules and 902 peptides. BBBP-Atlas achieved an accuracy of 0.8914 and an MCC of 0.7678 on the independent test set. On a balanced external benchmark of 200 compounds, our model reached an AUC of 0.9108, an accuracy of 0.8500, and an MCC of 0.7000, outperforming LightBBB by an absolute MCC gain of 6%. Case studies further showed that BBBP-Atlas captured clinically meaningful BBB permeability patterns, correctly identifying lorlatinib as BBB-permeable and vancomycin as BBB-impermeable with high confidence. The OmniBBBP-backed BBBP-Atlas offers a versatile and cross-modal approach for single-compound prediction, batch screening, and dataset exploration for CNS drug discovery. BBBP-Atlas is available at https://cadd.drugflow.com/bbbp/.

bioinformatics↗

bioRxiv2026

IgGM2: An All-Atom Foundation Model for Adaptive Immune Receptor Design

Accurate immune receptor design requires modeling the coupled variation of aminoacid sequence, full-atom conformation, and target-binding geometry across antibodies, nanobodies, and T-cell receptors (TCRs). Existing methods often address only part of this problem, either by separating structure generation from sequence design, relying on fixed-backbone inverse folding, or focusing on a single receptor class. We introduce IgGM2, a unified all-atom generative framework for immune receptor structure prediction and CDR sequence-structure co-design. IgGM2 follows a structure-to-design strategy: it first learns how immune receptors are positioned around fixed target structures, and then transfers this target-conditioned structural prior to CDR design. Unlike modular design pipelines, IgGM2 jointly generates CDR residue identities and full-atom receptor structures, allowing frame-work geometry to adapt to designed CDRs without separate inverse folding or external sidechain packing. Unlike continuous residue encodings based on virtualatom geometry, IgGM2 keeps sequence prediction explicit while using atom14 placeholders only for full-atom representation. On structure prediction benchmarks, IgGM2 better captures receptor-target spatial relationships than AlphaFold3 on FoldBench and achieves strong performance on TCR-pMHC modeling. On sequence design benchmarks, IgGM2 achieves competitive amino-acid recovery and improves Rosetta-based interface preference metrics, suggesting more favorable generated binding interfaces. These results support IgGM2 as a unified all-atom framework for adaptive immune receptor structure prediction and design.

bioinformatics↗

bioRxiv2026

Light-dependent and predator inducible aldehyde synthesis in Prochlorococcus for specific defense against Uronema

Cyanobacteria as a primary producer provide energy and carbon sources for the marine food web, of which predation-interactions play central roles in regulating global element cycles and marine ecosystem stability. Here, we report the anti-predation activity of a typical marine cyanobacterium Prochlorococcus MED4 to defend the predation by Uronema marinum. MED4 synthesize formaldehyde as the anti-predation chemical, of which the synthesis was light-dependent and predator-inducible. Compared to other protists, both the higher concentration of accumulated formaldehyde in U. marinum and the lower formaldehyde tolerance of U. marinum resulted in the specific anti-predation of MED4 against U. marinum. This specific anti-predation could regulate the cyanobacterial growth and the U. marinum infection of marine fishes. Metadata analyses showed the mutually exclusion of Prochlorococcus and Uronema in global marine environments. These findings significantly advance our understanding of the marine food web and biogeochemical cycles. Significance StatementPredation-driven interactions in the ocean are critical regulators of global biogeochemical cycles, yet active defense mechanisms in marine picophytoplankton remain largely unknown. This study reveals that Prochlorococcus MED4 as the most abundant primary producer in the ocean synthesize light-driven and predator-inducible aldehydes to actively and specifically defend predation against the ciliate Uronema marinum. The anti-predation of Prochlorococcus against Uronema could have broad implications in biocontrol of Uronema infection in marine fishes and regulation of the marine food web.

microbiology↗

bioRxiv2026

SCD-dependent lipid metabolism licenses alternative macrophage activation and macrophage plasticity

Lipid metabolic reprogramming accompanies macrophage activation, yet our understanding of why macrophages profoundly reshape their lipid composition remains unclear. Here, we identify stearoyl-CoA desaturase (SCD) as a lipid-metabolic checkpoint required for the acquisition of the alternatively activated macrophage (AAM) cell state. We show that SCD maintains lipid desaturation balance by ensuring the conversion of newly synthesized saturated long-fatty acids (SFAs) into monounsaturated fatty acids (MUFAs). Critically, disruption of SCD resulted in the rapid accumulation of saturated phosphatidic acid (PA) and hyperactivation of mTORC1, which perturbed the acquisition of stable AAM epigenetic and transcriptional programs. Notably, inhibition of upstream lipogenic enzymes (e.g., ACC and FASN) did not impair AAM polarization, underscoring that an imbalance between newly synthesized saturated and monounsaturated fatty acids, not the loss of de novo lipogenesis, determines AAM identity. Unexpectedly, the loss of SCD redirected IL-4-activated macrophages into an aberrant cell state poised for pro-inflammatory responses, indicating that disruption of this lipid metabolic checkpoint leads to misinterpretation of anti-inflammatory instructional cues. Rescue of SCD-deficient macrophages via mTORC1 inhibition confirmed a causal role for the SCD-PA-mTORC1 axis in controlling macrophage plasticity. Finally, we find that SCD deficiency and resultant PA accumulation attenuated Toxoplasma gondii-driven AAM polarization and reduced pathogen burden. Together, these data establish the SCD-PA lipid signaling axis as a critical metabolic regulator of macrophage plasticity with direct consequences for host defense.

immunology↗

bioRxiv2026

BioClaw: Human-Bot Research Collaboration Ecosystems in Group Chats

Day-to-day research discussions in group chats often generate hypotheses, analysis requests, and interpretation decisions, yet executing those analyses still requires researchers to leave the conversation and rely on fragmented local tools, databases, visualization software, and literature search engines. In this work, we present BioClaw, a human-bot research collaboration ecosystem that converts natural-language requests in group conversations into tool-grounded analyses executed within isolated Docker containers. Deployed across 8 messaging platforms, BioClaw turns each group chat into a persistent execution workspace. Its design combines multi-channel orchestration, per-group state and workspace management, and isolated containerized execution for reliable shared use over long-lived conversations. To support practical research workflows, BioClaw combines containerized execution with preinstalled 31 biomedical tools and 95+ skills. The application of BioClaw spans various biomedical domains (e.g., genomics, clinics, structural biology) and data modalities (e.g., sequencing data, EHR data, protein structure data). These results establish the viability of embedding executable, tool-rich agent workflows within shared digital workspaces, positioning group chats as a transformative paradigm for collaborative scientific discovery and innovation.

bioinformatics↗

bioRxiv2026

Integrative Multi-cohort Transcriptomics and Network Pharmacology Analysis Reveals Key Network Nodes and Potential Drug Clues in PCOS Granulosa Cells

BackgroundPolycystic ovary syndrome (PCOS) is a prevalent endocrine disorder with complex pathophysiology and limited therapeutic options. Identifying key molecular drivers and potential drug candidates is critical for improving clinical outcomes. MethodsWe integrated multi-cohort transcriptomics (GSE155489, GSE138518, GSE226146) with weighted gene co-expression network analysis (WGCNA), protein-protein interaction (PPI) network analysis, and drug repurposing. Differential expression analysis identified 1,039 DEGs, and WGCNA identified 10 PCOS-associated modules. Intersection of DEGs with module genes yielded 498 core candidate genes, which were subjected to functional enrichment, PPI network analysis, and connectivity map-based drug repurposing (CLUE/LINCS). Candidate drugs were further evaluated by molecular docking and ADMET prediction using a triple intersection strategy (hub genes, high differential expression, drug-target evidence). ResultsFunctional enrichment revealed significant enrichment in cell adhesion and TGF-beta signaling. PPI network analysis identified CD44 as the top hub gene (degree=42). Drug repurposing identified 106 candidate drugs, including troglitazone and enzalutamide. Using the triple intersection strategy, five genes (ID2, NR4A1, GJA5, ID1, MYH11) were prioritized for molecular docking. GJA5 showed strong predicted binding affinity with flufenamic acid (-7.88 kcal/mol), and cytosporone B exhibited favorable druglikeness (0 Lipinski violations). ConclusionThis study systematically characterizes PCOS-associated gene networks and provides a prioritized set of candidate targets and drugs through a purely computational framework. CD44 emerges as a key network node with potential relevance in PCOS pathophysiology. These findings offer testable hypotheses for future mechanistic studies and drug discovery efforts in PCOS.

systems biology↗

bioRxiv2026

ProtFlow: Flow Matching-based Protein Sequence Design with Comprehensive Protein Semantic Distribution Learning and High-quality Generation

Designing protein sequences with desired properties is a fundamental task in protein engineering. Recent advances in deep generative models have greatly accelerated this design process. However, most existing models face the issue of distribution centralization and focus on local compositional statistics of natural sequences instead of the global semantic organization of protein space, which confines their generation to specific regions of the distribution. These problems are amplified for functional proteins, whose sequence patterns strongly correlate with semantic representations and exhibit a long-tailed functional distribution, causing existing models to miss semantic regions associated with rare but essential functions. Here, we propose ProtFlow, a generative model designed for comprehensive semantic distribution learning of protein sequences, enabling high-quality sequence generation. ProtFlow employs a rectified flow matching algorithm to efficiently capture the underlying semantic distribution of the protein design manifold, and introduces a reflow technique enabling one-step sequence generation. We construct a semantic integration network to reorganize the representation space of large protein language models, facilitating stable and compact incorporation of global protein semantics. We pretrain ProtFlow on 2.6M peptide sequences and fine-tune it on antimicrobial peptides (AMPs), a representative class of therapeutic proteins exhibiting unevenly distributed activities across pathogen targets. Experiments show that ProtFlow outperforms state-of-the-art methods in generating high-quality peptides, and AMPs with desirable activity profiles across a range of pathogens, particularly against underrepresented bacterial species. These results demonstrate ProtFlows effectiveness in capturing the full training distribution and its potential as a general framework for computational protein design.

bioinformatics↗

bioRxiv2026

Ophiuchus-Ab: A Versatile Generative Foundation Model for Advanced Antibody-Based Immunotherapy

Antibodies exhibit extraordinary specificity and diversity in antigen recognition and have become a central class of therapeutics across a wide range of diseases. Despite this clinical success, antibody design remains fundamentally challenging. Antibody function emerges from intricate and highly coupled interactions between heavy and light chains, which complicate sequence-function relationships and limit the rational design of developable antibodies. Here, we reveal that modeling antibody sequence space at the level of paired heavy and light chains is essential to faithfully capture inter-chain dependencies, enabling a deeper understanding of antibody function and facilitating antibody discovery. We present Ophiuchus-Ab, a generative foundation model pre-trained on largescale paired antibody repertoires within a diffusion language modeling framework, unifying antibody generation and representation learning in a single probabilistic formulation. This framework excels diverse antibody design tasks, including CDR infilling, antibody humanization, and light-chain pairing. Beyond generation, diffusion-based pre-training yields transferable representations that enable accurate prediction of antibody properties, including developability, binding affinity, and specificity, even in low-data regimes. Together, these results establish Ophiuchus-Ab as a versatile foundation model for modeling antibodies, providing a foundation for next-generation antibody-based immunotherapy.

bioinformatics↗

bioRxiv2025

Inhibition mechanism of pancreatic KATP channels by centipede toxins

The pancreatic ATP-sensitive potassium (KATP) channel acts as a crucial metabolic sensor by regulating insulin secretion to maintain whole-body energy homeostasis. Gain-of-functional mutations in this channel lead to neonatal diabetes mellitus, a rare disorder in which certain mutants demonstrate resistance to standard sulfonylurea therapy. Recent studies have identified a centipede toxin, SpTx1, as a potent inhibitor of both human pancreatic KATP channels and their gain-of-functional mutants. This toxin stimulates insulin secretion, offering a promising therapeutic strategy for neonatal diabetes. Nevertheless, the molecular mechanism by which SpTx1 inhibits the KATP channel has remained elusive. Here, we report the crystal structure of SpTx1 and the cryo-electron microscopy (cryo-EM) structure of the KATP channel in complex with SpTx1. Our results demonstrate that SpTx1 binds to the extracellular surface of the Kir6.2 pore-forming subunit of the KATP channel. Multiple interactions at the SpTx1-Kir6.2 interface underpin the high affinity and specificity of SpTx1 for human Kir6.2. We show that SpTx1 inhibits potassium currents by physically blocking the ion conduction pore. Furthermore, a structure-guided search identified another centipede toxin, Sm3a, as a novel KATP channel blocker. Together, these findings provide key insights into the inhibitory mechanism of centipede toxins against the human Kir6.2-containing KATP channel and establish a foundation for developing these toxins into potential therapeutics targeting KATP channel-related diseases.

biophysics↗

bioRxiv2025

Off-target effects of PLX5622 revealed: mixing microglial function in anesthesia and addiction withdrawal

Microglia have gained increasing attention as important regulators in the brain. Acute microglial depletion using the colony-stimulating factor 1 receptor (CSF1R) inhibitor, PLX5622, has become a popular approach to uncover microglial function. PLX5622 treatment has been recently shown to significantly promote anesthetic emergence. In characterizing underlying mechanisms, we found that PLX5622 treatment did not change the global neuronal activity during anesthesia, but robustly enhanced anesthetic metabolism. Moreover, PLX5622 dramatically induced the expression of hepatic enzymes that extensively modify xenobiotic and endobiotic metabolism. Blocking increased enzymatic activity significantly reduced the arousal-promoting effects of PLX5622 in anesthesia, irrespective of microglial elimination. In addition, we demonstrate that enhanced drug metabolism also contributed to PLX5622-induced attenuation in nicotine-withdrawal anxiety. By revealing previously unrecognized effects of PLX5622, our findings raise caution in interpreting data generated from PLX5622 treatment and bring forward the need for designing more specific CSF1R inhibitors.

neuroscience↗

bioRxiv2025

Tissue-Specific Iron Levels Modulate Lipid Peroxidation and the FLASH Radiotherapy Effect

Iron is vital to living cells, playing a key role in cellular respiration, DNA synthesis, and various metabolic functions. Importantly, cancer cells have a higher dependency on iron compared to normal cells to support their rapid growth and survival. Due to this fact, tumors are more vulnerable to ferroptosis, an iron-dependent form of regulated cell death. Radiation therapy (RT), a standard treatment for many cancer patients, is known to induce ferroptosis. Ultra-high dose rate FLASH RT offers an improved therapeutic window by minimizing damage to normal tissues while preserving tumor control. However, the precise biological mechanisms behind the protective effects of FLASH RT on normal tissues remain unclear. In this study, we propose that variations in lipid peroxidation and ferroptosis, driven by intrinsic differences in iron levels between normal and cancerous tissues, contribute to this effect. Our findings show that FLASH RT increases lipid peroxidation and induces ferroptosis in tumor cells but does not significantly elevate lipid peroxidation and ferroptosis in normal tissues compared to conventional RT. To determine whether raising iron levels in normal tissues could abrogate the protective effects of FLASH, mice were fed a high-iron diet before RT. A high-iron diet before and after RT reversed the protective effect of FLASH, resulting in increased intestinal damage and lipid peroxidation. This suggests that baseline iron levels and iron-driven lipid peroxidation are critical factors in mediating the protective outcomes of FLASH RT. Overall, our study sheds light on the role of iron in modulating RT responses and provides new mechanistic insights into how FLASH RT influences normal and cancerous tissues.

cancer biology↗

bioRxiv2025

EpiMII: Integrating Structure and Graph Neural Networks for MHC-II Epitope and Neoantigen Design

MHC-II neoantigens play a critical role in immunotherapy, either as direct effectors or through their influence on CD8+ T cells. However, only a small fraction of tumor DNA mutations qualify as functional neoantigens, and current prediction tools often lack accuracy, leading to the low immunogenicity of predicted neoantigens in vivo. Here, we present EpiMII, a Graph Neural Network model for MHC-II epitope design, which learns from the structural features of epitopes to predict their sequences. To train EpiMII, we constructed a reliable, large dataset containing 142,934 MHC-II epitope structures. This approach achieves a 4.2x improvement over ProteinMPNN, with a sequence recovery rate of 78.0% for known MHC-II epitopes in the Protein Data Bank. As a case study, we designed a neoantigen from hepatocellular carcinoma. All five designed epitopes significantly activated CD4+ T cells in vitro and induced secretion of IFN-{gamma} and TNF-. Notably, one epitope treatment significantly reduced tumor volume in mice in vivo. EpiMII offers a novel and efficient approach for identifying MHC-II epitopes/neoantigens, potentially contributing to vaccine development.

immunology↗

bioRxiv2025

PepPCBench is a Comprehensive Benchmark for Protein-Peptide Complex Structure Prediction with AlphaFold3

Peptides mediate up to 40% of protein-protein interactions (PPIs), offering high specificity and the ability to target binding sites inaccessible to small molecules, making them promising candidates for drug development. Accurate modeling of protein-peptide complexes is crucial for understanding fundamental biological processes and for advancing peptide-based drug design. However, due to the high conformational flexibility of peptides, predicting their interactions with proteins remains a significant challenge. Recent advancements in artificial intelligence (AI), exemplified by all-atom protein folding neural networks (PFNNs) such as AlphaFold3 (AF3), have expanded predictive capabilities beyond proteins to encompass protein-peptide complexes. Nevertheless, existing evaluations of these methods are often limited in scope and lack systematic, multi-dimensional, and fair comparisons of PFNN performance in protein-peptide complex prediction. Here, we introduce PepPCBench, a comprehensive benchmark framework specifically developed to evaluate AF3s capabilities in predicting protein-peptide complexes. This study utilizes a carefully curated dataset named PepPCSet, which is excluded from the AF3s training or validation sets. This dataset includes 261 protein-peptide complexes with peptide lengths ranging from 5 to 30 residues. Our benchmark results indicate that AF3 outperforms other PFNNs in prediction accuracy and structural validation. However, its performance remains insufficient for practical peptide drug discovery, indicating room for improvement. It is expected that PepPCBench can provide some valuable insights into the enhancement of protein-peptide complex structure prediction and the development of peptide-based therapeutics. The dataset and pipeline protocols are available at https://github.com/zhaisilong/PepPCBench.

bioinformatics↗

bioRxiv2025

C9orf50 is a Targetable Spliceosome Regulator for Cancer Immunotherapy

Cancer cells harbor cell-intrinsic programs capable of sculpting immunogenic landscapes, yet molecular brakes that enforce immune evasion remain poorly defined. Here, we identify C9ORF50, a previously uncharacterized intrinsically disordered protein, as a liquid-liquid phase separation-driven regulator of RNA processing and actionable immunotherapy target. Genetic ablation of C9ORF50 induces selective intron retention in spliceosome component transcripts, leading to cytoplasmic accumulation of immunogenic double-stranded RNA (dsRNA). This dsRNA activates type I interferon pathway that in turn orchestrates a chemokine secretory program critical for T cell recruitment. The resultant enhancement of tumor immunogenicity and reprogramming of tumor immune microenvironment drives robust antitumor immunity, effectively converting immunologically "cold" tumors to "hot" phenotypes. Therapeutic targeting of C9ORF50 via RNA interference achieved robust tumor suppression in syngeneic models, suggesting its translational relevance. Notably, C9ORF50 exhibits tumor-predominant expression, with minimal baseline levels detected in healthy tissues or immune cell populations, underscoring its potential safety as a therapeutic target. Our study not only identifies C9ORF50 as a novel LLPS-dependent regulator of RNA splicing but also establishes it as a promising and potentially safe therapeutic target for cancer immunotherapy.

cancer biology↗

bioRxiv2025

Repurposing AZD-5991 for inhibiting growth and biofilm formation of Staphylococcus aureus by disrupting the cell membrane and targeting FabI.

Staphylococcus aureus infections have already become a major threat to public health worldwide. Drug resistance and biofilm formation are two critical factors that significantly affect the efficacy of the antimicrobial treatment of S. aureus infections using conventional antibiotics. Therefore, the discovery of novel antimicrobial agents with potent antibacterial and antibiofilm activity has become a hotspot in recent years. Here, we first demonstrated the remarkable inhibitory activity of AZD-5991, a selective Mcl-1 inhibitor, against S. aureus. The MIC50 and MIC90 of AZD-5991 against S. aureus were 12.5 M, and AZD-5991 significantly inhibited the growth of S. aureus at a subinhibitory concentration of 1/2 x MIC. Besides, AZD-5991 displayed bactericidal activity and a robust capacity for inhibiting biofilm formation against S. aureus with very low drug toxicity against host cell lines. Our data demonstrated the disruption of S.aureus cell integrity by AZD-5991 through membrane permeability assays and the ingredients of the bacterial phospholipids could neutralize the antibacterial activity of AZD-5991. Moreover, whole-genome sequencing and proteomic analysis were also applied to gain insights into the possible impact of AZD-5991 on the fatty metabolism of S. aureus. Furthermore, the antibacterial activity of AZD-5991 was significantly elevated by exogenous fatty acids linoleic acid (C18:2 {Delta} 9,12) and arachidonic acid (C20:4 {Delta} 5,8,11,14). Lastly, the biolayer interferometry assay supported the direct interaction of AZD-5991 with FabI, a necessary protein related closely to bacterial growth and fatty acid metabolism. Conclusively, this study suggests that AZD-5991 could inhibit the planktonic growth and biofilm formation of S. aureus by disrupting the cell membrane and targeting FabI. AZD-5991 might be a promising new antibiotic candidate for the antimicrobial treatment of S. aureus infections resistant to traditional clinical drugs. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=84 SRC="FIGDIR/small/638150v1_ufig1.gif" ALT="Figure 1"> View larger version (18K): org.highwire.dtl.DTLVardef@658444org.highwire.dtl.DTLVardef@cea949org.highwire.dtl.DTLVardef@1cc003dorg.highwire.dtl.DTLVardef@23de22_HPS_FORMAT_FIGEXP M_FIG C_FIG

microbiology↗

bioRxiv2024

A Multi-Modal Deep Learning Framework with Both Sequence and Structure for Tumor Antigens Prediction

Tumor antigens are key targets in cancer immunotherapies that can be recognized by T cell receptor and induce immune responses. However, precision screening of immunogenic tumor antigens remains a great challenge due to human leukocyte antigen (HLA) restriction and tumor antigen escape. Here, we introduce MultiTAP (Multi-modal Tumor Antigen Predictor), a pioneering multi-modal framework with TCR-peptide-HLA sequence and structure features incorporating an attention mechanism designed to accurately identify tumor antigens with immunogenic properties. By constructing the multi-modal TCR-peptide-HLA Dataset (TPHD) and integrating its sequence and structure, we perform antigen feature enhancement using peptide-HLA (pHLA) structural features at the residue level, achieving interpretable prediction of immunogenicity for tumor antigens. Relative to existing baseline models, the MultiTAP framework has exhibited superior efficacy in predicting the immunogenicity of tumor antigens. Through comprehensive out-of-distribution (OOD) assessments, MultiTAP has maintained predictive robustness across diverse HLA phenotypes and the continuously evolving landscape of epitope distributions. Overall, MultiTAP presents a brand-new and promising approach for cancer immunotherapies that target tumor antigens.

biophysics↗

bioRxiv2024

Calcium Homeostasis Modulator 2 Constitutes an ATP-regulation Pore in Mitochondria

Recent structural analyses showed that the calcium homeostasis modulator-2 (CALHM2) forms a mega channel, but its cellular location and endogenous function are yet unknown. We found that native CALHM2 resides on the mitochondrial inner membrane and constitutes an ATP-regulated ATP release channel. CALHM2 knockdown/knockout decreases cytosolic ATP concentration, and thereby compromises energy-sensitive processes, such as intracellular Ca2+ handling. However, CALHM2 loss-of-function elevates ATP concentration in the mitochondrial matrix, dephosphorylates key enzymes in the mammalian target of rapamycin (mTOR) pathway, and promotes longevity in CALHM2 knockout mice. These findings reveal that CALHM2 constitutes a novel regulator of mitochondrial metabolism, which may have important implications in aging and diseases.

cell biology↗