Biology subjects

Zhang, W.

Publications and source records attributed to Zhang, W..

At least 289 records · Page 16Linked to original sources

bioRxiv2025

Longitudinal analysis of retinal cell state transitions in RB1-deficient retinal organoids reveals the nascent cone precursors are the earliest cell-origin of human retinoblastoma

All cancers develop from malignant transformation of normal cells, while it is challenging to identify these "cells-of-origin" since it is impossible to witness the dynamic changes in human cancers. Retinoblastoma (Rb), an intraocular malignant cancer, is a typical model for study the molecular and cellular mechanisms of cancers. Although the maturing cone precursors (CPs) are proposed as the cellular origin of human Rb, it is unknow whether other retinal cells are also sensitive to RB1 inactivation. Here we developed RB1-deficient human retinal organoids (ROs) models from RB1-/- or RB1+/- human induced pluripotent stem cells (hiPSCs). RB1-/- hiPSCs developed into Rb tumors which recapitulated the features of human native Rb and had the ability to form consecutive orthotopic xenografts in mice. Importantly, RB1 loss induced the overproliferation of ATOH7+ neurogenic retinal progenitor cells (nRPCs), which disrupted the retinal development by generating proliferative, nascent retinal cells including retinal ganglion cells and CPs. scRNA-seq analysis verified the ATOH7+/RXR{gamma}+ nascent CPs survived and finally drove Rb development. In contrast, monoallelic RB1 inactivation with low pRB expression did not induce nascent CPs proliferation, but only induced nRPCs overproliferation which caused retinocytoma-like phenotype. Finally, a potential therapeutic agent for Rb was identified from multi-omics data. Our findings firstly indicate that nRPCs are the most sensitive cells to RB1 loss inducing nascent retinal cells abnormal proliferation, and ATOH7+ nascent CPs are the earliest cellular origin of human Rb, facilitating drug development for Rb.

cell biology↗

bioRxiv2025

Salt-inducible kinase inhibition promotes weight loss and improves the diastolic function of obesity-related heart failure with preserved ejection fraction in mice

BackgroundObesity negatively impacts cardiac function and is closely associated with heart failure with preserved ejection fraction (HFpEF). Salt-inducible kinases (SIKs) are critical regulators of energy metabolism and cardiovascular function. Small molecule SIK inhibitors have been developed but their effect in treating obesity-related HFpEF remains unexplored. We recently discovered that pharmacological SIK inhibition promotes the adipose tissue thermogenesis and mitochondrial biogenesis gene program. We reason that targeting SIKs to treat obesity-related complications would be beneficial for HFpEF. MethodsWe employed a preclinical HFpEF mouse model induced by two-hit stress of high-fat diet (HFD) and nitric oxide synthase (NOS) inhibition using L-NAME in drinking water. Eight-week-old C57BL/6J mice received a regular low-fat chow diet or HFD/L-NAME for 5 weeks and were treated with vehicle or a pan-SIK inhibitor YKL-05-099 (YKL) via daily intraperitoneal injection in the last 4 weeks. Body weight and parameters of adiposity, energy balance and glucose tolerance were assessed. Cardiovascular function was characterized by echocardiography and in vivo pressure-volume loop hemodynamic analysis. Myocardial transcriptomic data were analyzed to determine if changes in SIK gene expression are associated with human HFpEF. ResultsYKL treatment limits body weight gain mainly by reducing the fat mass in obese HFpEF mice. YKL-treated mice show better glucose tolerance, enhanced adipose tissue browning and decreased lipid deposition. YKL-treatment mice demonstrate preserved left ventricular (LV) ejection fraction, reduced LV filling pressure and improved diastolic function. Myocardial expression of SIK1, SIK2, and SIK3 mRNA is down-regulated in patients with HFpEF. However, higher SIK mRNA expression is associated with a subgroup of HFpEF patients that has a greater risk for HF hospitalization and/or death. ConclusionsTaken together, our study reveals a pathological role for SIKs in obesity-related HFpEF and suggests that pharmacological SIK inhibition would be a disease-modifying strategy for obese HFpEF, for which evidence-based therapy has been limited.

physiology↗

bioRxiv2025

MSMCE: A Novel Representation Module for Classification of Raw Mass Spectrometry Data

Mass spectrometry (MS) analysis plays a crucial role in the biomedical field; however, the high dimensionality and complexity of MS data pose significant challenges for feature extraction and classification. Deep learning has become a dominant approach in data analysis, and while some deep learning methods have achieved progress in MS classification, their feature representation capabilities remain limited. Most existing methods rely on single-channel representations, which struggle to effectively capture the structural information within MS data. To address these limitations, we propose a Multi-Channel Embedding Representation Module (MSMCE), which focuses on modeling inter-channel dependencies to generate multi-channel representations of raw MS data. Additionally, we introduce a "residual" connection along the channel dimension, significantly enhancing the classification performance of subsequent models. Experimental results on four public datasets demonstrate that the proposed MSMCE module not only achieves substantial improvements in classification performance but also reduces computational resource consumption and enhances training efficiency, highlighting its effectiveness and generalizability in raw MS data classification.

bioengineering↗

bioRxiv2025

An Integrative Multi-Omics Random Forest Framework for Robust Biomarker Discovery

High-throughput technologies now produce a wide array of omics data, from genomic and transcriptomic profiles to epigenomic and proteomic measurements. Integrating these diverse data types can yield deeper insights into the biological mechanisms driving complex traits and diseases. Yet, extracting key shared biomarkers from multiple data layers remains a major challenge. We present a multivariate random forest (MRF)-based framework enhanced by a novel inverse minimal depth (IMD) metric for integrative variable selection. By assigning response variables to tree nodes and employing IMD to rank predictors, our approach efficiently identifies essential features across different omics types, even when confronted with high-dimensionality and noise. Through extensive simulations and analyses of multi-omics datasets from The Cancer Genome Atlas, we demonstrate that our method outperforms established integrative techniques in uncovering biologically meaningful biomarkers and pathways. Our findings show that selected biomarkers not only correlate with known regulatory and signaling networks but can also stratify patient subgroups with distinct clinical outcomes. The methods scalable, interpretable, and user-friendly implementation ensures broad applicability to a range of research questions. This MRF-based framework advances robust biomarker discovery and integrative multi-omics analyses, accelerating the translation of complex molecular data into tangible biological and clinical insights.

bioinformatics↗

bioRxiv2025

Inwardly rectifying potassium channels regulate membrane potential polarization and direction sensing during neutrophil chemotaxis

Potassium channels regulate membrane potential and diverse physiological processes, including cell migration. However, the specific function of the inwardly rectifying potassium channels in immune cell chemotaxis is unknown. Here, we identified that the inwardly rectifying potassium channel Kir7.1 (KCNJ13) maintains the resting membrane potential and is required for directional sensing during neutrophil chemotaxis. Pharmacological or genetic inhibition of Kir7.1 in neutrophils impaired direction sensing toward various chemoattractants without affecting cell polarization in multiple neutrophil models. Using genetically encoded voltage indicators, we observed oscillating depolarization of the membrane potential in protrusions in zebrafish neutrophils, and Kir7.1 is required for polarized depolarization towards the chemokine source. Focal depolarization with optogenetic tools biases pseudopod selection and induces de novo protrusions. Global hyperpolarizing neutrophils stalled cell migration. Furthermore, Kir7.1 regulates GPCR signaling activation. This work adds membrane potential to the intricate feedforward mechanism, coupling the adaptive and excitable network required to steer immune cells in complex tissue environments.

cell biology↗

bioRxiv2025

DMBT1 promotes SARS-CoV-2 infection and its SRCR-derived peptide inhibits SARS-CoV-2 infection

DMBT1 is a large scavenger receptor cysteine rich (SRCR) B protein that has been reported as a tumor suppressor gene and a co-receptor for HIV-1 infection. Here we found DMBT1 is a major mucosal protein bound to SARS-CoV-2. Overexpression of DMBT1 in 293T cells enhanced infection by SARS-CoV-2 in ACE2 dependent manner. Blocking experiments using overlapping peptide library of SRCR domain of DMBT1 showed that CQGRVEVLYRGSWGTV peptide, which contains bacteria-binding VEVLXXXXW motif, could inhibit SARS-CoV-2 infection. High concentration of the peptide can significantly inhibit the replication of SARS-CoV-2 in hamsters. Single cell sequencing analysis showed the highest expression abundance of DMBT1 at airway submucosal glands. These results demonstrate that membrane DMBT1 can promote SARS-CoV-2 infection, and serve as a candidate target for antiviral development. Author SummaryWe discovered that a protein named as DMBT1, which is found in our mucous membranes, plays a significant role in SARS-CoV-2 infection. DMBT1 is a large glycoprotein that has previously reported linked to tumor suppression and HIV-1 infection. Our single-cell sequencing analysis revealed that DMBT1 is most abundant in the airway submucosal glands. In our study, we found that when DMBT1 is overexpressed in cells, it enhances SARS-CoV-2 infection. Using a peptide library from DMBT1, we pinpointed a specific peptide, CQGRVEVLYRGSWGTV, which can inhibit SARS-CoV-2 infection. At high concentrations, it significantly reduces the viruss ability to replicate in hamsters. These findings suggest that DMBT1 on the cell membrane can promote SARS-CoV-2 infection and highlight DMBT1 as a potential target for developing antiviral treatments.

microbiology↗

bioRxiv2025

Distinct allosteric networks in CDK4 and CDK6 in the cell cycle and in drug resistance

Cyclin-dependent kinases 4 and 6 (CDK4 and CDK6) are key regulators of the G1-S phase transition in the cell cycle. In cancer cells, CDK6 overexpression often outcompetes CDK4 in driving cell cycle progression, contributing to resistance against CDK4/6 inhibitors (CDK4/6i). This suggests distinct functional and conformational differences between these two kinases, despite their striking structural and sequence similarities. Understanding the mechanisms that differentiate CDK4 and CDK6 is crucial, as resistance to CDK4/6i--frequently linked to CDK6 overexpression--remains a significant therapeutic challenge. Notably, CDK6 is often upregulated in CDK4/6i-resistant cancers and rapidly proliferating hematopoietic stem cells, underscoring its unique regulatory roles. We hypothesize that their distinct conformational dynamics explain their differences in phosphorylation of retinoblastoma protein, Rb, inhibitor efficacy, and cell cycle control. This leads us to question how their dissimilar conformational dynamics encode their distinct actions. To elucidate their differential activities, molecular mechanisms, and inhibitor binding, we combine biochemical assays and molecular dynamics (MD) simulations. We discover that CDK4 and CDK6 have distinct allosteric networks connecting the {beta}3-C loop and the G-loop. CDK6 exhibits stronger coupling and shorter path lengths between these regions, resulting in higher kinase activity upon cyclin binding and impacting inhibitor specificity. We also discover an unrecognized role of the unstructured CDK6 C-terminus, which allosterically connects and stabilizes the R-spine, facilitating slightly higher activity. Our findings bridge the gap between the structural similarity and functional divergence of CDK4 and CDK6, advancing the understanding of kinase regulation in cancer biology. Graphical abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=107 SRC="FIGDIR/small/640857v2_ufig1.gif" ALT="Figure 1"> View larger version (33K): org.highwire.dtl.DTLVardef@cbdc16org.highwire.dtl.DTLVardef@1de54deorg.highwire.dtl.DTLVardef@195ed04org.highwire.dtl.DTLVardef@1fbc860_HPS_FORMAT_FIGEXP M_FIG C_FIG

molecular biology↗

bioRxiv2025

Generative and integrative modeling for transcriptomics with formalin fixed paraffin embedded material

Formalin-fixed paraffin embedded (FFPE) samples are challenging to profile using existing high-throughput sequencing technologies, including RNA-seq. This difficulty primarily arises from the degradation of nucleic acids, a problem that becomes particularly acute with samples stored for extended periods. FFPE-derived RNA-seq (fRNA-seq) data have a high rate of transcript dropout, a property shared with single cell RNA-seq. Transcript counts also have high variance and are prone to extreme values. We introduce the PaRaffin Embedded Formalin-FixEd Cleaning Tool (PREFFECT), a probabilistic framework for the analysis of fRNA-seq data. PREFFECT uses generative models to fit distributions to observed expression counts while adjusting for technical and biological variables. The framework can exploit multiple expression profiles generated from matched tissues for a single sample (e.g., a tumor and morphologically normal tissue) in order to stabilize profiles and impute missing counts. PREFFECT can also leverage sample-sample adjacency networks that assist graph attention mechanisms to identify the most informative correlations in the data. We demonstrate how PREFFECT uses this information to more accurately compute downstream analyses such as sample clustering in several datasets.

bioinformatics↗

bioRxiv2025

Unveiling Candidate Drivers in Cancer Progression Using Somatic-IV Analysis

MotivationIdentifying causal drivers of cancer progression is crucial for developing effective anti-cancer therapies. However, disentangling causality from correlation remains challenging due to confounding factors within the complex genetic and transcriptomic landscapes of cancer. ResultsTo address this challenge, we introduce the Somatic Instrumental Variable analysis (Somatic-IV), which integrates genetic, transcriptomic, and clinical outcome data to identify candidate driver genes likely playing a causal role in disease progression. Somatic-IV estimates genetic-exposure and genetic-outcome associations, utilizing MR-Egger regression to estimate bias-reduced causal effects. We assessed the statistical properties and performance of Somatic-IV through a simulation study and demonstrated its utility in pancreatic adenocarcinoma (PDAC) and colorectal cancer (CRC). Our analysis identified ZBED2 and CSNK2A2 as likely drivers of PDAC and CRC cancer progression, respectively. Somatic-IV is a novel approach for generating target hypotheses and guiding the development of effective therapies for cancer patients. Availability and implementationThe Somatic-IV package is available at: https://github.com/wzhang1984/Somatic_IV_package

bioinformatics↗

bioRxiv2025

Lactate metabolic coupling between the endplates and nucleus pulposus via MCT1 is essential for intervertebral disc health

During skeletal growth, there is an increased secretion of lactate by glycolytic nucleus pulposus (NP) cells of the intervertebral disc. To investigate the role of this anion, we generated annulus fibrosus (AF) and endplate (EP) specific Mct1cKO (Slc16a1Col2CreERT2) mice. Histological and spatial transcriptomic studies indicated significant disc degeneration in Mct1cKO, characterized by NP cell loss and delayed EP maturation. Metabolic assays showed that while AF and EP cells were glycolytic, EP chondrocytes readily metabolized lactate. In EP cells, lactate promoted protein, and H3K18 lactylation, implying epigenetic programming. These findings suggest that NP-derived lactate promotes EP cartilage transdifferentiation into the subchondral bone, and in its absence, continued glucose consumption by the persistent EP cartilage reduces glucose availability to the NP and AF likely contributing to tissue degeneration. This study provides the first in vivo evidence that metabolic coupling between NP and EP cells is essential for disc growth and health.

cell biology↗

bioRxiv2025

IMPDH Cytoophidia in the Pancreas: Morphological Dynamics and Metabolic Implications

Cytoophidia are filamentous structures formed by metabolic enzymes, playing critical roles in cellular metabolism and regulation. Inosine monophosphate dehydrogenase (IMPDH), the rate-limiting enzyme in de novo nucleotide synthesis, is known to form cytoophidia and is implicated in cell proliferation, nutrient sensing, and metabolic stress responses. Despite their significance, the behavior of IMPDH cytoophidia under different metabolic conditions, particularly in diabetes, remains poorly understood. This study investigates the distribution, morphology, and assembly patterns of IMPDH cytoophidia in the pancreas of normal and diabetic models. Using immunofluorescence staining, pharmacological interventions (e.g., mycophenolic acid), and glucose tolerance tests, we demonstrate that IMPDH cytoophidia exhibit distinct morphological changes in pancreatic islets under diabetic conditions. Notably, cytoophidia elongation is associated with impaired glucose tolerance, suggesting a potential link between cytoophidia dynamics and metabolic dysfunction. Our findings provide novel insights into the role of IMPDH cytoophidia in metabolic regulation and highlight their potential as therapeutic targets for diabetes and related metabolic disorders.

cell biology↗

bioRxiv2025

Pooled screening identifies combinatorial CAR signaling domains for next-generation CAR-M immunotherapies

Chimeric antigen receptor-engineered macrophages (CAR-Ms) hold great promise for solid tumor immunotherapy. The intracellular domains (ICDs) of CARs determine the phenotypic output of therapeutic macrophages but remain largely unexplored. Here, we constructed a CAR library containing 131 unique signaling domains derived from native immune receptors and identified 17 ICDs that enhance macrophage phagocytosis, inflammatory responses, or tumor infiltration in vitro and in vivo. We further developed a scalable 3 barcode technology, CARode, to uniquely label and trace ICD variants within large-scale combinatorial CAR library and applied it to single-cell RNA sequencing and single-cell CAR analysis to assess the synergetic effects of ICD combinations on macrophage activation. Our approach uncovered a novel CD40-LY9-FCRL1 chimeric receptor that modulates the tumor microenvironment and improves solid tumor clearance. In conclusion, our findings demonstrate that pooled screening can accelerate the discovery of complex ICD constructs, providing a powerful platform for engineering macrophage-based immunotherapies.

bioengineering↗

bioRxiv2025

Integrating multi-modal transcriptomics identifies cellular subtypes with distinct roles in PDAC progression

BackgroundPancreatic ductal adenocarcinoma (PDAC) features a complex tumor microenvironment (TME) that significantly influences patient outcomes. Understanding the TMEs cellular composition and interactions is crucial for identifying therapeutic targets to improve treatment. MethodsWe performed an integrative analysis combining 88 single-cell RNA sequencing (scRNA-seq) samples with 187,520 cells, 20 Visium spatial transcriptomics (ST-seq) samples with 67,933 spots, and 1383 bulk RNA-seq samples and 2 Xenium high-resolution ST-seq samples with 307,679 cells, to delineate and characterize distinct subpopulations of fibroblasts, macrophages, T/NK cells, and B/plasma cells within the PDAC microenvironment. Correlations among major cell types across 12 PDAC datasets were assessed through gene set variation analysis (GSVA). Kaplan-Meier survival analysis was used to evaluate the prognostic significance of cell subtypes. Furthermore, multiplex immunohistochemistry (mIHC) and the Xenium platform was utilized to validate cellular interactions in human PDAC tissues at both protein and RNA levels. ResultsSix fibroblast subtypes and eight macrophage/monocyte subtypes were identified. POSTN high fibroblasts and SPP1 high macrophages highly infiltrated tumor tissues and were associated with poor prognosis. Most immune cell subtypes mediate adverse prognoses, except for CCL4 high CD8+ T EFF cells and IGHG1 high plasma cells, which are linked to favorable outcomes. ST-seq revealed spatial colocalization of POSTN high Fibro and SPP1 high Macro cells, as well as CCL4 high CD8+ T EFF and IGHG1 high plasma cells. These findings were corroborated by mIHC and validated using Xenium spatial transcriptomics with single-cell resolution, which confirmed the expression and spatial proximity of these markers at both the gene and protein levels. ConclusionThis integrative analysis of the PDAC TME underscores the prognostic importance of POSTN high fibroblasts and SPP1 high macrophages, while also highlighting the protective roles of CCL4 high CD8+ T EFF cells and IGHG1 high plasma cells. These insights could improve PDAC treatment strategies.

cancer biology↗

bioRxiv2025

Construction of a low-temperature degrading synthetic bacterial community of corn straw

The stability of the cellulose structure in straw creates a challenge for short-term degradation in the natural environment, especially in the cold regions of north China. Therefore, to secure strains that can degrade straw at low temperatures, we selected soil samples from straw-returning land and used cellulose as the only carbon source. After conducting preliminary screening and further re-screening, we ultimately selected three strains that exhibited efficient degradation of corn straw at a temperature of 4 . Subsequently, the synthetic bacterial community C was constructed using these strains: Stenotrophomonas maltophilia, Flavobacterium johnsoniae and Pantoea rodasii. We conducted a 45-day solid-state and liquid-state fermentation experiment at low-temperature (12 ). The activity of lignocellulase was determined, and the enzyme production conditions were optimized. The results show that the liquid fermentation degradation rate of straw increased by 40 % compared with the control group in the natural state (without adding C). Similarly, the solid-state fermentation degradation rate of straw increased by 30 %, which was significantly faster than that of the control group. Finally, the enzyme production conditions of strain C were optimized as follows: culture time 96 h, inoculation amount 4 %, temperature 12 , pH 6.0, and Alkaline Carboxymethyl Cellulase (CMCase) activity 24.34 U/mL. The microbial system C efficiently degrades straw at low temperature, which is valuable for improving the comprehensive utilization rate of straw and has a broad application prospect in improving soil fertility.

microbiology↗

bioRxiv2025

Rprot-Vec: A deep learning approach for fast protein structure similarity calculation

The prediction of structural similarity and homology detection of protein sequences has always been a very challenging and valuable research. Based on protein structure similarity, rapid and accurate detection of homologous proteins can infer the functions of newly discovered or unannotated proteins. If we can use a method to quickly predict the structural similarity between two proteins, we can achieve the task of homologous detection of protein sequences. By using this method with partially known protein structures, we can quickly infer the functions of more unknown proteins. In this article, we propose an enhanced model Rprot-vec (Rapid Protein Vector) based on deep learning, which improves existing methods. It can quickly calculate the structural similarity of proteins and perform protein homology detection using only protein sequences. Meanwhile, our model can also be used to simulate proteins to accomplish more downstream tasks related to proteins. Our model performs well within the very important TM-score range, with an accurate similarity prediction rate of 65.3% for homologous proteins. In addition, for all protein pairs within the TM-score range, the average prediction error is 0.0561, indicating good performance. The superiority of our model can be demonstrated from various aspects. In addition, we have also supplemented the dataset in this field, providing data support for future researchers studying this field simultaneously. Contactzhangyichuan09@gmail.com Supplementary informationhttps://github.com/SuperZyccc/RProt-vec/

bioinformatics↗

bioRxiv2025

Single-Cell Genomics Elucidates Molecular Variations and Regulatory Mechanisms in Circulating Immune Cells

The human peripheral blood displays diverse molecular characteristics across populations, understanding the drivers and underlying mechanisms of which remains challenging. Here, we introduce the Chinese Immune Multi-Omics Atlas (CIMA), elucidating sex-, age-, and genetic-related molecular variations by analyzing multi-omics data from 428 adults with over 10 million immune cells. CIMA generated an enhancer-driven gene regulatory network, identifying 237 high-quality regulons and revealing cell type-specific regulatory mechanisms. Additionally, 11,521 lead cis-expression quantitative trait loci (eQTLs) and 46,339 chromatin accessibility QTLs (caQTLs) were identified at cell type level. CIMA also uncovered pleiotropic associations among immune-related disease risk loci, eQTLs, and caQTLs in a cell type-specific manner. Lastly, a novel cell language model, CIMA-CLM, was developed to predict chromatin accessibility and noncoding variant effects using chromatin sequences and gene expressions. This work represents a population-scale multi-omics resource of human immune cells, providing a valuable reference for future investigation of immune-related diseases.

immunology↗

bioRxiv2025

Local traveling waves of cytosolic calcium elicited by defense signals or wounding are propagated by distinct mechanisms

Cytosolic Ca2+ signatures with specific spatiotemporal patterns play crucial roles in plant responses to biotic and abiotic stresses. Perception of microbe- or damage-associated molecular patterns (MAMPs or DAMPs) initiates signaling cascades that represent the first layer of plant defense against pathogens known as pattern-triggered immunity (PTI). During PTI, MAMP/DAMP-induced cytosolic Ca2+ fluxes serve as essential messengers in the initiation and transmission of defense signals at the cellular, whole organ, and systemic levels. However, the specific patterns of these Ca2+ signatures in response to different pathogen cues, and the mechanisms that encode them, remain largely unexplored. In this study, we quantitatively assessed Ca2+ signatures at the single-cell level as well as the local traveling Ca2+ waves induced by global treatment of Arabidopsis cotyledons with MAMPs or DAMPs. We demonstrated that MAMPs induced distinct local spatiotemporal Ca2+ responses in epidermal pavement cells, with Ca2+ traveling waves consistently initiated from a subset of cells and spreading in an approximately radial pattern. These local traveling waves propagated at a slow but constant speed of approximately 1 {micro}m/s and spread to a limited number of neighboring cells. In contrast, wound-induced traveling waves displayed a diffusion-like decay pattern that moved rapidly away from the wounded cell but with diminishing speed over time and distance. Mathematical modeling supported a calcium-induced calcium release mechanism that could recapitulate the constant wave speed induced by MAMPs. These findings contribute to a deeper understanding of plant defense-related Ca2+ signaling mechanisms as well as how defense responses are spatially restricted within tissues. One Sentence SummaryA slow, local traveling wave of calcium is initiated from a subset of epidermal cells during plant immune signaling.

cell biology↗

bioRxiv2025

Reg-GPTTM: A Conversational AI Model for Enhanced Decision-Making in Regenerative Medicine

BackgroundAs artificial intelligence (AI) continues to transform various aspects of our lives, conversational AI models have become increasingly sophisticated. The development of more accurate and informative language processing assistants has significant implications for numerous fields, including health care, medical service, and research assistance. Materials and MethodsReg-GPT was developed by the Maharaj Institute of Immune Regenerative Medicine (MIIRM) using a combination of supervised and unsupervised learning techniques. The LLaMa 3.1 models parameters were fine-tuned using vast amounts of text data, enabling Reg-GPT to learn from its interactions with users. ResultsOur evaluation shows that Reg-GPT model performs well in several key areas, including response accuracy, fluency, and engagement. The results highlight the potential benefits of integrating Reg-GPT into regenerative medicine (RM) applications. ConclusionThis article provides a comprehensive introduction to Reg-GPT, showcasing its capabilities, performance, and potential uses. We believe that Reg-GPT has the potential to provide significant value in the RM and Medicare fields.

bioinformatics↗