Biology subjects

Zhang, L.

Publications and source records attributed to Zhang, L..

At least 541 records · Page 30Linked to original sources

bioRxiv2024

DSA-DeepFM: A Dual-Stage Attention-Enhanced DeepFM Model for Predicting Anticancer Synergistic Drug Combinations

MotivationDrug combinations are crucial in combating drug resistance, reducing toxicity, and improving therapeutic outcomes in the treatment of complex diseases. As the number of available drugs grows, the potential combinations increase exponentially, making it impractical to rely solely on biological experiments to identify synergistic drug pairs. Consequently, machine learning methods are increasingly used to systematically screen for synergistic drug combinations. However, most current approaches prioritize predictive performance by integrating auxiliary information or increasing model complexity. By overlooking the biological mechanisms behind feature interactions, their effectiveness in predicting drug synergy can be limited. ResultsWe present DSA-DeepFM, a deep learning model that integrates a dual-stage attention (DSA) mechanism with Factorization Machines (FMs) to improve drug synergy prediction by addressing complex biological feature interactions. The model incorporates categorical and auxiliary numerical inputs, embedding them into high-dimensional spaces and then fusing them through the DSA mechanism to capture both field-aware and embedding-aware patterns. These patterns are then processed by a DeepFM module, which captures low-order and high-order feature interactions before making final predictions. Validation testing demonstrates that DSA-DeepFM significantly outperforms traditional machine learning and state-of-the-art deep learning models. Additionally, t-SNE visualizations confirm the models discriminative power at various stages. As a case study, we use our model to identify eight novel synergistic drug combinations, three of which are well-supported by existing wet-lab experiments, underscoring its practical utility and potential for future applications.

bioinformatics↗

bioRxiv2024

RCANE: A Deep Learning Algorithm for Whole-genome Pan-Cancer Somatic Copy Number Aberration Prediction using RNA-seq Data

Transcriptome sequencing (RNA-seq) is widely used in cancer research to study the transcriptome and its role in disease progression. Somatic copy number aberrations (SCNAs) are key drivers of cancer development, and inferring SCNAs from RNA-seq data can provide critical insights for disease classification and treatment prediction. We introduce RCANE, a deep learning-based method designed to predict genome-wide SCNAs across various cancer types using RNA-seq data. RCANE is trained on data from The Cancer Genome Atlas (TCGA) and DepMap cancer cell lines, demonstrating superior performance compared to existing methods. This scalable approach offers a robust solution for improving SCNA prediction in cancer diagnostics and treatment.

bioinformatics↗

bioRxiv2024

UnCOT-AD: Unpaired Cross-Omics Translation Enables Multi-Omics Integration for Alzheimer's Disease Prediction

Alzheimers Disease (AD) is a progressive neurodegenerative disorder, posing a growing public health challenge. Traditional machine learning models for AD prediction have relied on single omics data or phenotypic assessments, limiting their ability to capture the diseases molecular complexity and resulting in poor performance. Recent advances in high-throughput multi-omics have provided deeper biological insights. However, due to the scarcity of paired omics datasets, existing multi-omics AD prediction models rely on unpaired omics data, where different omics profiles are combined without being derived from the same biological sample, leading to biologically less meaningful pairings and causing less accurate predictions. To address these issues, we propose UnCOT-AD, a novel deep learning framework for Unpaired Cross-Omics Translation enabling effective multi-omics integration for AD prediction. Our method introduces the first-ever cross-omics translation model trained on unpaired omics datasets, using two coupled Variational Autoencoders and a novel cycle consistency mechanism to ensure accurate bidirectional translation between omics types. We integrate adversarial training to ensure that the generated omics profiles are biologically realistic. Moreover, we employ contrastive learning to capture the disease specific patterns in latent space to make the cross-omics translation more accurate and biologically relevant. We rigorously validate UnCOT-AD on both cross-omics translation and AD prediction tasks. Results show that UnCOT-AD empowers multi-omics based AD prediction by combining real omics profiles with corresponding omics profiles generated by our cross-omics translation module and achieves state-of-the-art performance in accuracy and robustness. Source code is available at https://github.com/abrarrahmanabir/UnCOT-AD

bioinformatics↗

bioRxiv2024

Microbial succession, community assembly and adaptation over five years in a newly discovered deep-sea cold seep

Deep-sea cold seeps, characterized by the release of hydrocarbon-rich fluids such as methane from the seafloor, create unique ecosystems with evolving biological communities. Microorganisms serve as pioneers, initially colonizing early-stage cold seeps and providing essential matter and energy for subsequent colonizers. However, microbial succession, community assembly, and adaptation in these environments remain largely unexplored due to challenges associated with locating early-stage cold seeps. Here, we collected sediment samples from a newly discovered early-stage cold seep site, QDN-W07, in the South China Sea at varying depths over five years. We also sampled two previously reported cold seep sites, QDN-S18 and HM-ROV01, for comparative analysis. Amplicon sequencing revealed a high relative abundance of ANME-3 (up to 99%) and Methylomonadaceae (up to 83%) in surface sediments, and indicated changes in microbial community structures over time. Dispersal limitation dominated community assembly, while the contribution of homogeneous selection decreased over time, accompanied by a declined in the intensity of microbial association. Metagenomic analysis of core microbiome, comprising 172 metagenomic-assembled genomes at QDN-W07, revealed tolerance to low temperatures and oxygen. Core microbiome genomes were enriched with genes related to aerobic methane oxidation, urea metabolism, and the oxidation of sulfur, ammonia and carbon monoxide. These results indicate that core microorganisms obtain energy from the oxidation of inorganic compounds, allowing them to thrive in cold, methane-rich and unstable conditions. Overall, this study enhances our understanding of deep biosphere life, biogeochemical cycling, and the ecological management of cold seep environments.

microbiology↗

bioRxiv2024

Ictal-Related Chirp as a Biomarker for Monitoring Seizure Progression

Despite being prevalent, the causes, mechanisms, and progression of epilepsy--a chronic neurological disorder with unprovoked seizures--are not well understood, complicating drug development for treatment. This study used a comprehensive mouse epilepsy kindling model dataset to investigate frequency modulation (chirp) as a potential indicator of distinct states of epilepsy (early evoked discharge, late evoked discharge, spontaneous recurrent seizure, and drug state). Employing time-frequency ridge extraction, chirp identification, and statistical testing, our analyses revealed that chirp patterns occur in the majority of ictal discharges (>81.6%), persisting across evoked and spontaneous seizures. While the focus was on hippocampal recordings, chirps were also detected in the piriform peripheral cortex. Significant frequency and duration changes in chirp patterns during the transition from early to late evoked ictal events suggest their potential as the screening tool for seizure progression. Additionally, detailed analyses illuminate the impact of Lorazepam, a GABAA enhancer, on chirp characteristics, providing insights into how increased inhibitory tone quantifiably influences excitatory-inhibitory balances during seizures.

neuroscience↗

bioRxiv2024

IGF2 mediates Hippo signaling to control liver size

The Hippo pathway is a central mechanism in organ size control, but the mediator for its function remains elusive. Here, we show that the expression of insulin-like growth factor II (IGF2) is directly induced by YAP/TAZ transcription cofactors of the Hippo pathway in a developmental stage- and cell type-specific manner. In mouse livers, Igf2 expression is sustained by YAP/TAZ in hepatoblasts and immature hepatocytes at fetal and neonatal stages, coupling with rapid cell proliferation and liver size expansion, whereas turned off in matured hepatocytes following YAP/TAZ inactivation to prevent liver overgrowth. In contrast, YAP/TAZ fails to regulate Igf2 expression in cholangiocytes and diverse liver mesenchymal cells where epigenetic barriers, including DNA and histone methylation, are implanted near Igf2 promoters. Furthermore, IGF2 activates IGF1R signaling and promotes normal and neoplastic liver growth, while the inactivation of IGF2 or IGF1R effectively blocks YAP/TAZ-induced hepatomegaly and hepatoblastoma. In conclusion, our findings reveal that IGF2 is a bona fide mediator of the Hippo pathway, indicating a fundamental role of the Hippo-IGF2-IGF1R signaling axis in organ size regulation and malignancies.

cell biology↗

bioRxiv2024

Deep structural analysis of rhamnogalacturonan-I (RG-I) in Arabidopsis reveals new insights into pectin diversity

Pectic polysaccharides are an integral part of primary plant cell walls, where they are perfrom crucial structural and biological functions. Pectin is generally divided into four distinct structural categories, including homogalacturonan, xylogalacturonan, rhamnogalacturonan I (RG-I) and rhamnogalacturonan II. Among the four main pectin domains, the structural intricacies of rhamnogalacturonan-I (RG-I) remain the least understood, especially in widely employed plant models. We employed multiple complementary analytical techniques to present a detailed structural analysis of RG-I in the model system Arabidopsis thaliana. Using highly purified RG-I from different tissues, we conducted comparative linkage and NMR analyses, complemented by mass spectrometry of enzymatically digested RG-I oligosaccharides. Our findings present the most comprehensive structural overview of Arabidopsis RG-I to date, revealing novel structural features. Notably, we identified O-acetylation of rhamnose backbone residues as a predominant feature, a modification previously unreported in this species. The combined results present a comprehensive structural overview of Arabidopsis thaliana RG-I that will serve as a roadmap for studying pectin biosynthesis and function.

biochemistry↗

bioRxiv2024

WRAD core perturbation impairs DNA replication fidelity promoting immunoediting in pancreatic cancer

It is unclear how cells counteract the potentially harmful effects of uncoordinated DNA replication in the context of oncogenic stress. Here, we identify the WRAD (WDR5/RBBP5/ASH2L/DPY30) core as a modulator of DNA replication in pancreatic ductal adenocarcinoma (PDAC) models. Molecular analyses demonstrated that the WRAD core interacts with the replisome complex, with disruption of DPY30 resulting in DNA re-replication, DNA damage, and chromosomal instability (CIN) without affecting cancer cell proliferation. Consequently, in immunocompetent models, DPY30 loss induced T cell infiltration and immune-mediated clearance of highly proliferating cancer cells with complex karyotypes, thus improving anti-tumor efficacy upon anti-PD-1 treatment. In PDAC patients, DPY30 expression was associated with high tumor grade, worse prognosis, and limited response to immune checkpoint blockade. Together, our findings indicate that the WRAD core sustains genome stability and suggest that low intratumor DPY30 levels may identify PDAC patients who will benefit from immune checkpoint inhibitors.

cancer biology↗

bioRxiv2024

40 Hz sensory stimulation enhances CA3-CA1 coordination and prospective coding during navigation in a mouse model of Alzheimer's disease

40 Hz sensory stimulation ("flicker") has emerged as a new technique to potentially mitigate pathology and improve cognition in mouse models of Alzheimers disease (AD) pathology. However, it remains unknown how 40 Hz flicker affects neural codes essential for memory. Accordingly, we investigate the effects of 40 Hz flicker on neural representations of experience in the hippocampus of the 5XFAD mouse model of AD by recording 1000s of neurons during a goal-directed spatial navigation task. We find that an hour of daily exposure to 40 Hz audio-visual stimulation over 8 days leads to higher coordination between hippocampal subregions CA3 and CA1 during navigation. Consistent with CA3s role in generating sequential activity that represents future positions, 40 Hz flicker exposure increased prospective coding of future positions. In turn, prospective coding was more prominent during efficient navigation behavior. Our findings show how 40 Hz flicker enhances key hippocampal activity during behavior that is important for memory. Significance StatementBrain stimulation has emerged as a new potential therapeutic approach to potentially correct or improve altered neural activity in Alzheimers disease. One such approach, 40 Hz sensory stimulation, or flicker, has been shown to improve cognition in disease models. However, it is not clear how 40 Hz flicker affects neural activity underlying memory processes. Here, we investigate how 40 Hz flicker exposure affects neural activity patterns that are crucial for memory. We find 40Hz flicker increases neural coordination in memory circuits, indicating better communication. Furthermore, 40Hz flicker increased neural representations of future positions, patterns theorized to support memory-based planning. These results indicate that 40 Hz flicker increases key neural activity that is important for memory.

neuroscience↗

bioRxiv2024

Connectomic Organization of the Suprachiasmatic Nucleus

The suprachiasmatic nucleus (SCN), the mammalian master clock, is a special structure dedicated to the time computation. However, a connectomic understanding of the nucleus as a whole is lacking. Using serial section electron microscopy, we reconstructed multiscale and multimodal SCN communication networks. Intra-SCN synaptic network consists of 9,566 morphologically similar neurons and 4.3 million synapses, organized into multiscale circuitries that interact with SCN-traversing axon fascicles. Strikingly, SCN neurons engage in unique soma-soma ephaptic interaction, forming 2,038 electrotonic integrative units with the largest overlapping the light-responsive area (LRA). SCNs paracrine modality contains 47,396 m3 dense core vesicles, with notable scarcity in the LRA, suggesting cross-modal coordination and functional integration. These distinct network features provide comprehensive insights into the system architecture of the mammalian circadian clock.

neuroscience↗

bioRxiv2024

S100a4+ alveolar macrophages accelerate the progression of precancerous atypical adenomatous hyperplasia by promoting fatty acid metabolism

Lung cancer is preceded by premalignant lesions, and what factors drive this transformation and the potential regulatory mode in the context of tumor initiation remain to be elucidated. In the course of precancerous lesions, we found a phasic shift in metabolic patterns. Macrophages are a heterogeneous cell population with high plasticity in the tumor microenvironment. Single-cell interaction and metabolic analyses highlighted a cellular state, S100a4+ alveolar macrophages, which exhibited distinct fatty acid metabolic activity, such as palmitic acid metabolism, at the atypical adenomatous hyperplasia stage, accompanied by an angiogenic-promoting function in a pre-neoplastic setting. These findings were reproducible in human single-cell transcriptomes and had been confirmed by histopathological staining and in vitro cell coculture assays. Taken together, the results from this study demonstrated that the S100a4+ alveolar macrophage subset contributes to tumorigenesis by altering its metabolic state, suggesting that metabolic interventions targeting this cell state in the early stage of disease may delay neoplastic transformation of the lung epithelium.

cancer biology↗

bioRxiv2024

Integrated temporal profiling of iPSCs-derived motor neurons from ALS patients carrying C9orf72, FUS, TARDBP, and SOD1 mutations

Amyotrophic Lateral Sclerosis (ALS) is a lethal neurodegenerative disease that damages motor neurons in the central nervous system, causing progressive muscle weakness that ultimately leads to death. However, its underlying mechanisms still need to be fully understood, particularly the heterogeneity and similarity between various gene mutants during disease progression. In this study, we conducted temporal RNA-seq profiling in human induced pluripotent stem cells (hiPSCs) and iPSC-derived motor neurons (iMNs) carrying the C9orf72, FUS, TARDBP, and SOD1 mutations from both ALS patients and healthy individuals. We discovered dysregulated gene expression and alternative splicing (AS) throughout iMN development and maturation, and ALS iMNs display enrichment of cytoskeletal defects and synaptic alterations from premature stage to mature iMNs. Our findings indicate that synaptic gene dysfunction is the common molecular hallmark of fALS, which might result in neuronal susceptibility and progressive motor neuron degeneration. Analysis of upstream splicing factors revealed that differentially expressed RNA-binding proteins (RBPs) in ALS iMNs may cause abnormal AS events, suggesting the importance of studying RBP defects in ALS research. Overall, our research provides a comprehensive and valuable resource for gaining insights into the shared mechanisms of ALS pathogenesis during motor neuron development and maturation in iMN models.

neuroscience↗

bioRxiv2024

Multi-omics investigation of spontaneous T2DM macaque reveals gut microbiota promote T2DM by up-regulating the absorption of excess palmitic acid

Although gut microbiota and lipid metabolites have been suggested to be closely associated with type 2 diabetes mellitus (T2DM), the interactions between gut microbiota, lipid metabolites and the host in T2DM development remains unclear. Rhesus macaques may be the best animal model to investigate these relationships given their spontaneous development of T2DM. We identified eight spontaneous T2DM macaques and conducted a comprehensive study investigating the relationships using multi-omics sequencing technology. Our results from 16S rRNA, metagenome, metabolome and transcriptome analyses identified that gut microbiota imbalance, tryptophan metabolism and fatty acid {beta} oxidation disorders, long-chain fatty acid (LCFA) accumulation, and inflammation occurred in T2DM macaques. We verified the accumulation of palmitic acid (PA) and activation of inflammation in T2DM macaques. Importantly, mice transplanted with spontaneous T2DM macaque fecal microbiota and fed a high PA diet developed prediabetes within 120 days. We determined that gut microbiota mediated the absorption of excess PA in the ileum, resulting in the accumulation of PA in the serum consequently leading to T2DM in mice. In particular, we demonstrated that the specific microbiota composition was probably involved in the process. This study provides new insight into interactions between microbiota and metabolites and confirms causative effect of gut microbiota on T2DM development.

microbiology↗

bioRxiv2024

DDX3 regulates the cap-independent translation of the Japanese encephalitis virus via its interactions with PABP1 and the untranslated regions of the viral genome

The translation of global cellular proteins is almost completely repressed in cells with flavivirus infection, while viral translation remains efficient. The mechanisms of flaviviruses evade host translational shutoff are largely unknown. Here, we identified viral elements and host factors associated with JEV evasion of host shutoff. JEV 5'UTR lacked IRES or IRES-like activity, while noncapped 5'UTR initiated translation in the presence of 3'UTR. Furthermore, the elements DB2 and sHP-SL within 3'UTR were involved in the regulation of cap-independent translation, which is conserved in the genus Orthoflavivirus. By RNA affinity purification and mass spectrometry analysis, cellular DDX3 and PABP1 were identified as key factors in regulating cap-independent translation of JEV via their interactions with DB2 and sHP-SL RNA structures. Mechanistically, we revealed that DDX3 could bind to both 5'UTR and 3'UTR of the JEV genome to establish a closed-loop architecture, recruit eIF4G/eIF4A to form the DDX3/PABP1/eIF4G/eIF4A tetrameric complex via its interaction with PABP1, thereby recruiting 43S PIC to the 5'-end of the JEV genome to start translation. Our findings demonstrated a noncanonical translation strategy employed by JEV and further revealed the regulatory roles of DDX3 and PABP1 in this mechanism. These results expand our knowledge of the translation initiation regulation in flaviviruses under the state of host translational shutoff, which provides a conserved antiviral target against orthoflavivirus.

microbiology↗

bioRxiv2024

General purpose genotypes and evolution of higher plasticity in clonality underlie knotweed invasion

Many widespread invasive plant species express high phenotypic variation across novel environments, providing a unique opportunity to examine ecological and evolutionary dynamics under global change. However, biogeographical studies often lack information about the origin of introduced populations, limiting our understanding of post-introduction evolution of introduced species in the new range. Here, we assessed the responses of Reynoutria japonica from 128 populations spanning three latitudinal transects in the native ranges of China and Japan, and the introduced ranges of North America and Europe. When grown in two common gardens in the native range, plants from introduced populations in North America and Europe differed in almost all traits from those from native Chinese populations, but were similar to plants from native Japanese populations. Compared to Chinese populations, North American, European and Japanese populations expressed lower trait values and plasticity in most traits. However, plants from both introduced ranges and from Japanese populations expressed higher clonality and plasticity in clonality than plants from Chinese populations. In addition, introduced populations expressed higher plasticity in clonality but lower plasticity in basal diameter compared to Japanese populations. Our study showed heritable differences in phenotypes between plants from the native ranges of Japan and China, and those from the introduced ranges were similar to the putative source of origin in Japan. However, we found that introduced populations may have evolved higher plasticity in clonal growth. Our findings emphasize the critical role of clonality and plasticity in invasion success, demonstrating the importance of discriminating between source and non-source native populations to identify ecological and evolutionary responses of invasive plants to novel environments.

ecology↗

bioRxiv2024

Synthesis and Evaluation of Carmofur Analogs as Antiproliferative Agents, Inhibitors to the Main Protease (Mpro) of SARS-CoV-2, and Membrane Rupture-Inducing Agents

Initially developed as a derivatized analog of 5-fluorouracil for the treatment of colorectal cancer, carmofur has more recently demonstrated potent covalent inhibition of the main protease (Mpro) of SARS-CoV-2. Harnessing our previously described workflow for the optimized preparation of carmofur using benchtop 19F NMR spectroscopy, here, we prepared and evaluated a synthetic library of nine carmofur analogs with a selection of side chain motifs or single-atom substitution to explore the diversifiability of these compounds as Mpro inhibitors, where we discovered that a hexyl carbamate analog outperformed carmofur, and as antiproliferative agents in model human cell lines to identify differences in potency when the carbonyl electrophilicity and/or alkyl side chains are modified. Finally, we describe a novel workflow for the evaluation of membrane-rupturing small molecules through imaging of fluorescently labeled giant unilamellar vesicles (GUVs), and through this, we identified two lipophilic urethane analogs of carmofur bearing dodecyl urethane and octadecyl urethane side chains that have potent membrane-rupturing capability in the nanomolar range, providing insight into a potential mechanism for the in vitro activities of lipidated 5-fluorouracil analogs.

biochemistry↗

bioRxiv2024

Induction of the ISR by AB5 subtilase cytotoxin drives IFN-I expression in pDCs via STING activation

We demonstrate that exposure to the AB5 subtilase cytotoxin (SubAB) induces the unfolded protein response (UPR) in human peripheral blood mononuclear cells, concomitant with a pro-inflammatory response across distinct cell subsets. Notably, SubAB selectively induces type-I interferon (IFN) expression in plasmacytoid dendritic cells, acting synergistically with Toll-like receptor 7 stimulation. The induction of type-I IFN in response to SubAB relies on stimulator of interferon genes (STING) activation, coupled with protein synthesis inhibition mediated by protein kinase R-like endoplasmic reticulum kinase and phosphorylation of the eukaryotic translation initiation factor 2 subunit-alpha. By impeding mRNA translation through the integrated stress response, SubAB precipitates the downregulation of the negative innate signaling feedback regulator Tax1-binding protein 1. This downregulation is necessary to unleash TANK-binding kinase 1 signaling associated with STING activation. These findings shed new light on how UPR-inducing conditions may regulate the immune system during infection or pathogenesis.

immunology↗

bioRxiv2024

Recursive Cleaning for Large-scale Protein Data via Multimodal Learning

AO_SCPLOWBSTRACTC_SCPLOWReliable datasets and high-performance models work together to drive significant advancements in protein representation learning in the era of Artificial Intelligence. The size of protein models and datasets has grown exponentially in recent years. However, the quality of protein knowledge and model training has suffered from the lack of accurate and efficient data annotation and cleaning methods. To address this challenge, we introduce ProtAC, which corrects large Protein datasets with a scalable Automatic Cleaning framework that leverages both sequence and functional information through multimodal learning. To fulfill data cleaning, we propose the Sequence-Annotation Matching (SAM) module in the model, which filters the functional annotations that are more suitable for the corresponding sequences. Our approach is a cyclic process consisting of three stages: first pretraining the model on a large noisy dataset, then finetuning the model on a small manually annotated dataset, and finally cleaning the noisy dataset using the finetuned model. Through multiple rounds of "train-finetune-clean" cycles, we observe progressive improvement in protein function prediction and sequenceannotation matching. As a result, we achieve (1) a state-of-the-art (SOTA) model that outperforms competitors with fewer than 100M parameters, evaluated on multiple function-related downstream tasks, and (2) a cleaned UniRef50 dataset containing [~]50M proteins with well-annotated functions. Performing extensive biological analysis on a cleaned protein dataset, we demonstrate that our model is able to understand the relationships between different functional annotations in proteins and that proposed functional annotation revisions are reasonable.

bioengineering↗