Biology subjects

Huo, M.

Publications and source records attributed to Huo, M..

14 recordsLinked to original sources

bioRxiv2026

DisenTE: Sparse Pattern-Context Modeling for Interpretable Translation-Efficiency Matrix Completion

Partially observed object-by-context matrices arise across data-rich science, where dominant object effects can obscure smaller but informative context-dependent variation. We study this problem in a translation-efficiency atlas of 9,494 5' UTRs across 78 cellular and tissue contexts. We present DisenTE, a sequence-conditioned neural model that combines separate sequence and context branches with a sparse low-rank pattern-context channel. Each module pairs a sequence-derived activation with context-specific deployment weights, forming a dictionary whose sequence and context components can be examined separately. Under five-fold within-panel entry masking, DisenTE achieves a UTR-centered residual Spearman correlation of 0.641 +/- 0.005, compared with 0.304 +/- 0.003 for the strongest reference model. The learned dictionary retains 11 of 20 candidate modules. CTM 6 has the largest overlap with an external TOP set and a cap-proximal pyrimidine pattern; CTMs 5 and 7 also overlap the set but have purine-containing consensuses. The evidence supports CTM 6 as a TOP sequence anchor and CTMs 5 and 7 as TOP-set-associated factors. On this dataset, DisenTE improves completion over the evaluated references and provides module-level summaries of its fitted context-dependent variation.

bioinformatics↗

bioRxiv2026

Allogeneic CRISPR-Engineered CAR-T Cells Drive Potent Antitumor Activity in Solid Tumors

Chimeric antigen receptor (CAR) T-cell therapy has shown limited efficacy in solid tumors, in part due to variability in autologous T cells derived from heavily pretreated patients with advanced disease. To address these constraints, we developed an off-the-shelf allogeneic CAR-T platform using CRISPR-Cas9-mediated genome editing in T cells from healthy donors to enable targeted CAR insertion at the TRAC locus with concurrent disruption of B2M. Using adeno-associated virus (AAV) delivery, we designed CAR-T cells targeting glypican-2 (GPC2) and glypican-3 (GPC3), emerging antigens expressed in pediatric and adult solid tumors. Genome-edited allogeneic CAR-T cells exhibited potent, antigen-specific cytotoxicity across multiple tumor models. GPC2-directed allogeneic CAR-T cells demonstrated enhanced or comparable activity relative to conventional lentiviral CAR-T cells in neuroblastoma models and mediated tumor regression with prolonged survival in preclinical models. Notably, repeated dosing augmented antitumor efficacy without evidence of toxicity, supporting multi-dose regimens for solid tumors. Similarly, GPC3-targeted allogeneic CAR-T cells based on a single-domain antibody showed robust activity against hepatocellular carcinoma cells in vitro and in vivo. These findings establish a scalable, genome-engineered allogeneic CAR-T strategy with strong therapeutic potential and support the clinical development of off-the-shelf cell therapies for pediatric and adult solid tumors.

immunology↗

bioRxiv2025

OmicsSankey: Crossing Reduction of Sankey Diagram on Omics Data

In bioinformatics, Sankey diagrams have been widely used to elucidate complex biological insights by visualizing gene expression patterns, microbial community dynamics, and cellular interactions. However, computational scalability remains a challenge for large-scale biological networks. In this work, we present OmicsSankey, a novel formulation of the layout optimization problem for Sankey Diagrams that employs eigen decomposition as a closed-form solution, addressing graph disconnection through a teleportation mechanism that enhances connectivity and stabilizes eigenvector solutions. Experimental results on synthetic datasets with varying layers and nodes validate the efficacy of OmicsSankey compared to state-of-the-art layout-optimizers. Improving the Sankey layouts for Cell Layers, BioSankey, and Sequence Flow further validates OmicsSankey in enhancing the interpretability of biological insights.

bioinformatics↗

bioRxiv2025

The first chromosomal-level genome assembly of an allotetraploid oyster

Tetraploid oysters are used to cross with diploids to produce triploid oysters that have become an important part of the oyster aquaculture industry worldwide. Although most tetraploid oysters are artificially induced autotetraploids, allotetraploids can be produced between closely related species, providing new opportunities for polyploid breeding and studying genome interactions. Using PacBio sequencing, Illumina sequencing, and high-throughput chromosome conformation capture scaffolding, we assembled the first high-quality genome assembly of an artificially induced allotetraploid between the Pacific oyster Crassostrea gigas and Portuguese oyster Crassostrea angulata. The assembled genome is 1.23 Gb, with a contig N50 of 2.56 Mb and a scaffold N50 of 57.22 Mb, and anchored to 20 chromosomes. The assembly contains 58,330 protein-coding genes, 98.34% of which are functionally annotated. The heterozygosity and the ratio of repetitive sequences is 5.50% and 46.43%, respectively. This chromosomal-level genome assembly of an allotetraploid oyster provides a valuable genetic resource for studying genome biology of molluscs and for advanced breeding of polyploids that are critical for the oyster aquaculture industry.

genomics↗

bioRxiv2025

TransHLA: A Hybrid Transformer Model for HLA-Presented Epitope Detection

BackgroundPrecise prediction of epitope presentation on human leukocyte antigen (HLA) molecules is crucial for advancing vaccine development and immunotherapy. Conventional HLA-peptide binding affinity prediction tools often focus on specific alleles and lack a universal approach for comprehensive HLA site analysis. This limitation hinders efficient filtering of invalid peptide segments. ResultsWe introduce TransHLA, a pioneering tool designed for epitope prediction across all HLA alleles, integrating Transformer and Residue CNN architectures. TransHLA utilizes the ESM2 large language model for sequence and structure embeddings, achieving high predictive accuracy. For HLA class I, it reaches an accuracy of 84.72% and an AUC of 91.95% on IEDB test data. For HLA class II, it achieves 79.94% accuracy and an AUC of 88.14%. Our case studies using datasets like CEDAR and VDJdb demonstrate that TransHLA surpasses existing models in specificity and sensitivity for identifying immunogenic epitopes and neoepitopes. ConclusionsTransHLA significantly enhances vaccine design and immunotherapy by efficiently identifying broadly reactive peptides. Our resources, including data and code, are publicly accessible at https://github.com/SkywalkerLuke/TransHLA Key PointsO_LIWe developed TransHLA, a deep learning tool for predicting epitopes across all HLA alleles using Transformer and Residue CNN architectures. C_LIO_LIThe model uses ESM2 embeddings to improve predictive accuracy and efficiency. C_LIO_LITransHLA shows superior specificity and sensitivity in identifying immunogenic epitopes and neoepitopes compared to existing models. C_LIO_LIOur approach offers potential advancements in vaccine design and immunotherapy through enhanced peptide analysis. C_LI

bioinformatics↗

bioRxiv2024

Multi-Modal Large Language Model Enables All-Purpose Prediction of Drug Mechanisms and Properties

Accurately predicting the mechanisms and properties of potential drug molecules is essential for advancing drug discovery. However, traditional methods often require the development of specialized models for each specific prediction task, resulting in inefficiencies in both model training and integration into work-flows. Moreover, these approaches are typically limited to predicting pharmaceutical attributes represented as discrete categories, and struggle with predicting complex attributes that are best described in free-form texts. To address these challenges, we introduce DrugChat, a multi-modal large language model (LLM) designed to provide comprehensive predictions of molecule mechanisms and properties within a unified framework. DrugChat analyzes the structure of an input molecule along with users queries to generate comprehensive, free-form predictions on drug indications, pharmacodynamics, and mechanisms of action. Moreover, DrugChat supports multi-turn dialogues with users, facilitating interactive and in-depth exploration of the same molecule. Our extensive evaluation, including assessments by human experts, demonstrates that DrugChat significantly outperforms GPT-4 and other leading LLMs in generating accurate free-form predictions, and exceeds state-of-the-art specialized prediction models.

pharmacology and toxicology↗

bioRxiv2024

Multi-Modal Large Language Model Enables Protein Function Prediction

Predicting the functions of proteins can greatly accelerate biological discovery and applications, where deep learning methods have recently shown great potential. However, these methods predominantly predict protein functions as discrete categories, which fails to capture the nuanced and complex nature of protein functions. Furthermore, existing methods require the development of separate models for each prediction task, a process that can be both resource-heavy and time-consuming. Here, we present ProteinChat, a versatile, multi-modal large language model that takes a proteins amino acid sequence as input and generates comprehensive narratives describing its function. ProteinChat is trained using over 1,500,000 (protein, prompt, answer) triplets curated from the Swiss-Prot dataset, covering diverse functions. This novel model can universally predict a wide range of protein functions, all within a single, unified framework. Furthermore, ProteinChat supports interactive dialogues with human users, allowing for iterative refinement of predictions and deeper exploration of protein functions. Our experimental results, evaluated through both human expert assessment and automated metrics, demonstrate that ProteinChat outperforms general-purpose LLMs like GPT-4, one of the flagship LLMs, by over ten-fold. In addition, ProteinChat exceeds or matches the performance of task-specific prediction models.

bioinformatics↗

bioRxiv2023

sRNAfrag: A pipeline and suite of tools to analyze fragmentation in small RNA sequencing data

Fragments derived from small RNAs such as small nucleolar RNAs hold biological relevance. However, they remain poorly understood, calling for more comprehensive methods for analysis. We developed sRNAfrag, a standardized workflow and set of scripts to quantify and analyze sRNA fragmentation of any biotype. In a benchmark, it is able to detect loci of mature microRNAs fragmented from precursors and, utilizing multi-mapping events, the conserved 5 seed sequence of miRNAs which we believe may extraoplate to other small RNA fragments. The tool detected 1411 snoRNA fragment conservation events between 2/4 eukaryotic species, providing the opportunity to explore motifs and fragmentation patterns not only within species, but between. Availability: https://github.com/kenminsoo/sRNAfrag.

bioinformatics↗

bioRxiv2023

Remodeling the central metabolism of Escherichia coli enables a universal chassis

E. coli is the host of choice to produce a wide variety of chemicals and proteins. Overflow metabolism is considered as the widespread and major obstacle in microbial synthesis, and overcoming this common bottleneck may enable a universal chassis. Here, we constructed an E. coli universal chassis (ABKS strain) with significantly suppressed overflow metabolism, presenting similar growth rate, decreased glucose consumption, and increased production of desired chemicals and proteins when compared with wild-type BL21(DE3) strain. Furthermore, we demonstrated that metabolic flux of ABKS strain was reprogrammed from TCA cycle to glyoxylate bypass at isocitrate node via the synergistic effect of multi-layer regulation in gene transcription and protein modification. This metabolic reconfiguration alleviates overflow metabolism, avoids CO2 release in TCA cycle, finally improving the carbon atom economy in bioprocess. Our chassis has widespread and practical use for elevating the production and yield of multiple desired chemicals and proteins from different carbon source. The metabolic reconfiguration also provides theoretical basis for rational design of efficient bioproduction strains.

bioengineering↗

bioRxiv2023

TCR2vec: a deep representation learning framework of T-cell receptor sequence and function

The T-cell receptor (TCR) repertoires are critical components of the adaptive immune system, and machine learning methods were proposed to analyze the TCR repertoire data. However, most methods work solely on the hypervariable CDR3 regions of TCRs, overlooking the information encoded in other domains. Representing full TCRs as informative vectors can be indispensable for developing reliable and effective machine learning models. We introduce TCR2vec, a deep representation learning framework with 12 layers of transformer blocks, to pave the way for downstream modelings of full TCRs. Together with masked language modeling (MLM), we propose a novel pretraining task named similarity preservation modeling (SPM) to capture the sequence similarities of TCRs. Through a multi-task pretraining procedure on MLM and SPM, TCR2vec learns a contextual understanding of TCRs within a similarity-preserved vector space. We first verify the effectiveness of TCR2vec in predicting TCRs binding specificity and TCR clustering through comparison with three other embedding approaches. TCR2vec can be finetuned on small task-specific labeled data for enhanced performance, which outperforms state-of-the-art models by 2-25% in predicting TCRs binding specificity. Next, we compare the performance of two versions of TCR2vec pretrained on full TCRs (TCR2vec) or CDR3s (CDR3vec) and demonstrate that TCR2vec consistently outperforms CDR3vec by 12-14%. Further analysis of attention maps reveals that residues outside CDR3 also make notable contributions to the recognition of antigens. TCR2vec is available at https://github.com/jiangdada1221/TCR2vec.

bioinformatics↗

bioRxiv2022

TEINet: a deep learning framework for prediction of TCR-epitope binding specificity

The adaptive immune response to foreign antigens is initiated by T-cell receptor (TCR) recognition on the antigens. Recent experimental advances have enabled the generation of a large amount of TCR data and their cognate antigenic targets, allowing machine learning models to predict the binding specificity of TCRs. In this work, we present TEINet, a deep learning framework that utilizes transfer learning to address this prediction problem. TEINet employs two separately trained encoders to transform TCR and epitope sequences into numerical vectors, which are subsequently fed into a fully connected neural network to predict their binding specificities. A major challenge for binding specificity prediction is the lack of a unified approach to sample negative data. Here, we first assess the current negative sampling approaches comprehensively and suggest that the Unified Epitope is the most suitable one. Subsequently, we compare TEINet with three baseline methods and observe that TEINet achieves an AUROC of 0.760, which outperforms baseline methods by 6.4-26%. Furthermore, we investigate the impacts of the pretraining step and notice that excessive pretraining can adversely affect model performance. Our results and analysis show that TEINet can make an accurate prediction using only the TCR sequence (CDR3{beta}) and the epitope sequence, providing novel insights to understand the interactions between TCRs and epitopes. TEINet is available at https://github.com/jiangdada1221/TEINet.

bioinformatics↗

bioRxiv2021

Gastric cancer derived exosomes induce peritoneal mesothelial cell EMT through TGF-β1/Smads pathway to promote peritoneal metastasis

Epithelial-mesenchymal transition (EMT) plays an important role in peritoneal metastasis of Gastric cancer (GC). Tumor exosomes can mediate tumor directed metastasis, and TGF-{beta}1 is an important factor in inducing tumor Epithelial mesenchymal transition. However, it is not clear whether GC derived exosomes can induce peritoneal mesothelial cells through the TGF-{beta}1/ Smads pathway and the effect of injured peritoneal mesothelial cells on the biological characteristics of GC cells. In this study, we demonstrated that GC-derived exosomes can activate the TGF-{beta}1/Smads pathway in peritoneal mesothelial cells and induce the corresponding EMT process, and that the injured peritoneal mesothelial cells can improve the migration and adhesion of GC cells. Taken together, these data further support the critical role of exosomes in the remodeling of the pre-metastatic microenvironment.

cancer biology↗

bioRxiv2021

Unbalanced dietary patterns contribute to the pathogenesis of precocious puberty by affecting gut microbiota and host metabolites

Precocious puberty (PP) mostly stems from endocrine disorders. However, its triggering factors, especially for the early onset of partial PP, and the associated pathogenic mechanisms remain ambiguous. In this study, a systematic analysis in the form of a questionnaire of lifestyles, gut microbiome, and serum metabolome data was carried out to examine the pathogenesis of PP in a cohort comprised of 200 girls, with or without PP. The analysis revealed substantial alterations in gut microbiota, serum metabolites, as well as lifestyle patterns in the PP group, which were characterized by an elevated abundance of {beta}-glucuronidase-producing and butyrate-producing bacteria, and excessive lipid concentration with decreased levels of organic nitrogen compounds in the serum of the participants. These differential microbes and metabolites tend to be reliable non-invasive diagnostic biomarkers aiding the early diagnosis of PP and exhibit a strong discriminative power (AUC = 0.93 and AUC = 0.97, respectively). Furthermore, the microbial biomarkers were confirmed in an independent validation cohort (n = 83, AUC = 0.85). Moreover, structural equation modeling revealed that unhealthy dietary habits were the primary contributors for the alteration of gut microbiota and serum metabolites, triggering the imbalance in the host hormones that leads to premature physical development. Our study determines a causal relationship among the gut microbiota, host metabolites, diet, and clinical characteristics of preadolescent girls who experienced early onset of PP, and formulates non-invasive diagnostic tools demonstrating excellent performance for the early detection of PP.

microbiology↗

bioRxiv2019

Nilotinib activates endothelial TLR4 to exacerbate atherosclerosis

The use of nilotinib (Tasigna(R)), a second-generation tyrosine kinase inhibitor for treating chronic myeloid leukemia, increases risks for atherosclerosis. Here, we demonstrate that in endothelial cells, nilotinib activated TLR4, triggerd expression of inflammatory molecules, and increased monocyte attachment, which were all inhibited by knockdown of TLR4 or TLR4 inhibitor, CLI-095. Orally administered nilotinib profoundly accelerated atherosclerotic lesion formation in ApoE-/- mice, while co-administration of CLI-095 effectively reduced lesion areas. Our findings reveal TLR4 activation as an underlying mechanism of the pro-atherosclerotic effect of nilotinib and suggest TLR4 inhibition as an effective therapeutic approach to address vascular safety issue of nilotinib.

cell biology↗