Biology subjects

Jie, H.

Publications and source records attributed to Jie, H..

8 recordsLinked to original sources

bioRxiv2026

Pan-Screening the Structural Predictability of Drug Toxicity: Validation of Molecular Fingerprints

Objective To systematically evaluate the predictability of drug chemical structures against the complete set of MedDRA Preferred Terms (PTs), construct a comprehensive structure adverse reaction landscape, validate the utility of molecular fingerprint representations in pan-screening tasks, and establish quantitative tools for drug safety pre-screening. Methods Drug adverse event association data were compiled from SIDER 4.1 and its derivative compilation Nsider, encompassing 2,005 approved drugs, of which 1,120 small-molecule drugs with successfully retrieved chemical structures were retained for analysis. For each Preferred Term (PT) meeting the minimum sample-size threshold ([≥]25 positive and [≥]30 negative drugs; n = 768), XGBoost classifiers were trained using MACCS 166-bit fingerprints, with model performance evaluated via 5-fold stratified cross-validation. A PT was designated as high-confidence if it achieved a validation AUC [≥] 0.75, a cross-validation AUC [≥] 0.70, and passed Benjamini Hochberg false discovery rate (FDR) correction (q < 0.05). The entire analytical pipeline was replicated using Morgan ECFP4 2048-bit fingerprints to assess the robustness of fingerprint selection. The robustness of significance screening was independently validated by permutation testing (500 random label permutations), and model generalizability was evaluated through Bemis-Murcko scaffold-based stratified split validation. SHAP (SHapley Additive exPlanations) analysis was employed to identify critical structural features, and mechanism-based case studies further demonstrated concordance between model-identified features and established pharmacophores. Results Among the 768 modelable PTs, 71 (9.2%) met the high-confidence significance criteria (FDR q < 0.05), predominantly comprising nervous system disorders (10 PTs) and psychiatric disorders (9 PTs). The toxicities exhibiting the strongest structural predictability included hyperprolactinaemia (AUC = 0.968), dermatitis acneiform (AUC = 0.966), tardive dyskinesia (AUC = 0.952), and breast tenderness (AUC = 0.979). Under a unified XGBoost hyperparameter configuration, the low-dimensional MACCS fingerprint (166-bit) demonstrated near-equivalent performance to the high-dimensional Morgan fingerprint (2048-bit) in pan-screening tasks ({Delta}AUC = -0.012, Cohen's d = 0.16), and achieved comparable performance at a substantially lower dimensionality across 18 of 22 organ systems. SHAP analysis identified key structural fragments corresponding to established pharmacophores (six-membered rings, tertiary amine cores, heteroatomic bonds, and bridged heteroatomic chains), and mechanism-based case studies further demonstrated that the model captured structural patterns concordant with known pharmacophores of D2 antagonists, EGFR inhibitors, and related agents. Bemis-Murcko scaffold-based stratified split validation revealed a median AUC decrease of merely 0.042, with 66% of PTs maintaining an AUC [≥] 0.70 under scaffold-split conditions. Immune-mediated and multifactorial endpoints remained structurally opaque under the current data conditions. Conclusion This study constructed a comprehensive structure-adverse reaction predictability landscape encompassing 768 MedDRA PTs; seventy-one PTs achieved high-confidence criteria. The MACCS 166-bit fingerprint achieved near-equivalent performance to the Morgan 2048-bit fingerprint at a markedly lower dimensionality, and its chemically explicit features enable direct translation of model outputs into structural alerts, providing a computational tool for prospective screening of potential adverse reaction profiles for any novel compound. A tiered assessment strategy is proposed: machine learning ensemble models for high-confidence screening, substructure alert rules for high-specificity exclusion, and pharmacovigilance signal detection for complementary coverage of structurally unpredictable endpoints. Bemis-Murcko scaffold-based stratified split validation further demonstrated that the median generalization performance decrease was merely 0.042 AUC, with 66% of PTs maintaining an AUC [≥] 0.70 under scaffold-split conditions, corroborating the existence of a genuine chemical basis underlying structure-adverse reaction mappings.

pharmacology and toxicology↗

bioRxiv2026

A reactivated thalamocortical plasticity window promotes learning and is reshaped by experience

Adult sensory loss can reactivate critical-period-like thalamocortical plasticity, but whether this reactivation defines a temporally gated circuit state that facilitates learning and is reciprocally shaped by experience remains unknown. Here we define its in vivo trajectory and functional consequences in adult mouse barrel cortex. Infraorbital nerve transection opened a transient window of enhanced layer 4 thalamocortical gain. Training during this window lowered whisker-detection thresholds and promoted learning by accelerating the transition to stable performance. Local GluN2B blockade prevented both cortical potentiation and the learning advantage, linking critical-period-associated plasticity mechanisms to adaptive behavior in the adult brain. Neuropixels recordings showed that weak inputs preferentially increased neuronal responses, whereas strong inputs produced sharper temporal coding. The relationship was reciprocal: experience reshaped the trajectory of this circuit state, with training before the normal peak advancing the emergence of potentiation, training during the active window prolonging the potentiated state, and training after closure failing to reinstate potentiation. State prolongation accompanied more persistent sensory memory. These findings establish a reciprocal, timing-dependent interaction between endogenous plasticity and experience, revealing a general principle by which adult circuits can convert transient plastic potential into adaptive behavioral change and informing strategies that align training with periods of heightened plasticity.

neuroscience↗

bioRxiv2026

Distinct melanoma EV subpopulations reflect immune- and stress-induced tumor states

Melanoma is the deadliest form of skin cancer, and improved non-invasive approaches to monitor tumor burden and immune dynamics are needed. Although extracellular vesicles (EVs) are increasingly explored as cancer biomarkers, how distinct EV subpopulations reflect dynamic tumor states induced by immune pressure remains insufficiently understood. Using proteomic analyses, we identified the melanoma-associated antigens gp100 (PMEL) and GPNMB in EVs derived from B16F10 melanoma cells and incorporated them into sandwich enzyme-linked immunosorbent assays (ELISAs) that capture total EVs while selectively detecting gp100 and GPNMB EV subpopulations. We subsequently evaluated these EV populations in murine models of anti-tumor vaccination and in plasma samples from melanoma patients. In vivo, melanoma-associated EV subpopulations increased in the serum of tumor-bearing mice and were further augmented following antigen-specific anti-tumor vaccination, whereas total CD81 EVs accumulated more gradually. Notably, gp100 EV levels, but not GPNMB EVs, correlated with tumor-infiltrating lymphocyte densities across multiple time points and treatment conditions. In vitro, TNF/IFN{gamma} stimulation preferentially increased total CD81 EV release, whereas gp100 EVs were promoted by IL-1{beta} stimulation. In contrast, GPNMB EVs accumulated more gradually and broadly across inflammatory, hypoxic, and cytotoxic stress conditions. Together, these findings indicate that immune and stress signals differentially remodel melanoma-associated EV composition. The assay translated to humans, revealing elevated gp100 EV levels in the plasma of melanoma patients compared with healthy donors, whereas GPNMB EVs identified a subset of melanoma patients. We describe a clinically feasible approach that enables direct detection of melanoma-associated EV subpopulations from blood without prior EV isolation. Conceptually, our findings suggest that immune and stress signals dynamically shape circulating EV composition, generating distinct EV signatures that reflect tumor state.

immunology↗

bioRxiv2024

Harnessing Smartphone RGB Imagery and LiDAR Point Cloud for Enhanced Leaf Nitrogen and Shoot Biomass Assessment - Chinese Spinach as a Case Study

Accurate estimation of leaf nitrogen concentration and shoot dry-weight biomass in leafy vegetables is crucial for crop yield management, stress assessment, and nutrient optimization in precision agriculture. However, obtaining this information often requires access to reliable plant physiological and biophysical data, which typically involves sophisticated equipment, such as high-resolution in-situ sensors and cameras. In contrast, smartphone-based sensing provides a cost-effective, manual alternative for gathering accurate plant data. In this study, we propose an innovative approach for estimating leaf nitrogen concentration and shoot biomass by integrating smartphone RGB imagery with Light Detection and Ranging (LiDAR) data, using Amaranthus dubius (Chinese spinach) as a case study. The influence of varying nitrogen dosages on individual spectral and structural features derived from smartphone RGB imagery and LiDAR data was modeled. Additionally, the spectral indices from RGB imagery and structural indices from LiDAR data were combined to model both leaf nitrogen concentration and shoot biomass. The performance of crop parameter modeling was evaluated using support vector regression, random forest regression, and lasso regression. Results demonstrate that the combined use of smartphone RGB imagery and LiDAR data can accurately estimate leaf total reduced nitrogen concentration, leaf nitrate concentration, and shoot dry-weight biomass, with average relative root mean square errors as low as 0.06, 0.16, and 0.05, respectively. Furthermore, the optimal nitrogen dosage for maximizing biomass yield in Chinese spinach was also estimated using the smartphone data. This study lays the groundwork for smartphone-based estimate leaf nitrogen concentration and shoot biomass, supporting accessible precision agriculture practices.

bioinformatics↗

bioRxiv2024

Age Dependent Integration of Cortical Progenitors Transplanted at the Adult CSF-Brain Interface

There has been renewed interest in neural transplantation of cells and tissues for brain repair. Recent studies have demonstrated the ability of transplanted neural precursor cells and in vitro grown organoids to mature and locally integrate into host brain neural circuitry. Much effort has focused on how the transplant behaves and functions after the procedure, but the extent to which the host brain can properly innervate the transplant, particularly in the context of aging, is largely unexplored. Here we report that transplantation of rat embryonic cortical precursor cells into the cerebrospinal fluid-subventricular zone (CSF-SVZ) of adult rat brains generates a brain-like tissue (BLT) at an ectopic site. This model allows for the assessment of long-range connectivity and cellular interactions between the transplant and the host brain as a function of host age. The transplanted precursor cells initially proliferate, then differentiate, and develop into mature BLTs, which receive supportive cellular components from the host including blood vessels, microglia, astrocytes, and oligodendrocytes. There was integration of the BLT into the host brain which occurred at all ages studied, suggesting that host age does not affect the maturation and integration of the transplant-derived BLT. Long-range axonal projections from the BLT into the host brain were robust throughout the different aged recipients. However, long-distance innervation originating from the host brain into the BLT significantly declined with age. This work demonstrates the feasibility and utility of integrating new neural tissue structures at ectopic sites into adult brain circuits to study host-transplant interactions.

neuroscience↗

bioRxiv2023

The role of RNA in the maintenance of chromatin domains as revealed by antibody mediated proximity labelling coupled to mass spectrometry

Eukaryotic chromatin is organised in individual domains, which are defined by their location within the nuclear space, the molecular interactions within each other and their distinct proteomic compositions. In contrast to cellular organelles surrounded by lipid membranes, the composition of distinct chromatin domains is rather ill described and highly dynamic. To identify proteins associated with a specific chromatin domain and to better understand how those domains are established and maintained, we used a new method, which we termed AMPL-MS (Antibody mediated proximity labelling mass spectrometry). This method is based on antibodies and does not require the expression of fusion proteins and therefore constitutes a versatile and very sensitive method to characterize chromatin domains containing specific signature proteins or histone modifications. We used AMPL-MS to characterize the composition of chromocenter as well as the chromosome territory containing the hyperactive X-chromosome in Drosophila. As an outcome of this method, we show the importance of RNA in maintaining the integrity of these domains as treatment with RNAse alters their proteomic composition. Our data demonstrate the power of AMPL-MS to characterize the composition of non-membranous nuclear domains.

biochemistry↗

bioRxiv2022

DNA binding domain undergoes dynamic and selective protein-protein interactions to facilitate CTCF insulation

CTCF is required for three-dimensional chromatin organization and a predominant insulator protein. However, its roles in insulating enhancers have not been fully explained in 3D nuclear organization. Here, we found that the CTCF DNA binding domain (DBD) forms dynamic self-interacting clusters. We next investigated the spatial relationships between these clusters and other transcription regulators with a light-induced imaging system. Strikingly, CTCF DBD clusters were found to incorporate other insulator proteins but are not coenriched with transcriptional activators in the nucleus. This property is not observed in other domains of CTCF or the DBDs of other transcription factors. Moreover, endogenous CTCF shows a phenotype consistent with the DBD by forming small protein clusters and interacts less transcriptional activators bound, CTCF motif arrays. Our results reveal an interesting phenomenon that CTCF DBD interacts with insulator proteins and selectively localizes to nuclear positions with lower concentrations of transcriptional activators, providing new insights into the insulation function of CTCF.

molecular biology↗

bioRxiv2019

An integrated Asian human SNV and indel benchmark combining multiple sequencing methods

Precision medicine of human requires an accurate and complete reference variant benchmark for different populations. A human standard cell line of NA12878 provides a good reference for part of the human populations, but it is still lack of a fine reference standard sample and variant benchmark for the Asians. Here, we constructed a stabilized cell line of a Chinese Han volunteer. We received about 4.16T clean data of the sample using eight sequencing strategies in different laboratories, including two BGI regular NGS platforms, three Illumina regular NGS platforms, two linked-read libraries, and PacBio CCS model. The sequencing depth and reference coverage of eight sequencing strategies have reached the saturation. We detected small variants of SNPs and Indels using the eight data sets and obtained eight variant sets by performing a series of strictly quality control. Finally, we got 3.35M SNPs and 349K indels supported by all of sequencing data, which could be considered as a high confidence standard small variant sets for the studies. Besides, we also detected 5,913 high quality SNPs located in the high homologous regions supported by both linked-reads and CCS data benefited by their long-range information, while these regions are recalcitrant to regular NGS data due to the limited mappability and read length. We compared the later SNPs against the public databases and 969 sites of them were novel SNPs, indicating these SNPs provide a vital complement for the variant database. Moreover, we also phased more than 99% heterozygous SNPs also supported by linked-reads and CCS data. This work provided an integrated Asians SNV and indel benchmark for the further basic studies and precision medicine.

genomics↗