Biology subjects

Zhan, F.

Publications and source records attributed to Zhan, F..

6 recordsLinked to original sources

bioRxiv2026

STCS: A Platform-Agnostic Framework for Cell-Level Reconstruction in Sequencing-Based Spatial Transcriptomics

Sequencing-based spatial transcriptomics technologies, including Visium HD and Stereo-seq, now enable transcriptome-wide profiling at subcellular resolution. However, these platforms generate measurements over spatially barcoded units rather than biologically segmented cells, creating a fundamental bottleneck for cell-centric analysis and interpretation. Robust recon-struction of coherent single-cell transcriptomes from high-density spatial bins remains an unresolved computational challenge. Here we present STCS (Spatial Transcriptomics Cell Segmentation), a platform-agnostic framework that reconstructs cell-level gene expression profiles by integrating nuclei segmentation with a joint transcriptomic-spatial distance model. STCS is governed by two interpretable parameters and incorporates a reference-free parameter selection strategy based on internal stability and spatial coherence metrics, enabling adaptable deployment across tissue types and technologies without requiring matched ground-truth annotations. We benchmark STCS on a Visium HD human lung cancer dataset with matched Xenium-derived cell segmentation, enabling direct cell-level validation, and on high-resolution Stereo-seq mouse brain data to assess cross-platform generalizability. Across multiple evaluation dimensions--including cell-type agreement, spatial organization, gene-expression fidelity, and compositional accuracy--STCS achieves consistent improvements over existing methods while preserving biologically coherent spatial structure. As sequencing-based spatial transcriptomics is rapidly adopted across biomedical research, STCS provides a broadly applicable and open-source solution for reconstructing cell-resolved transcriptomes, facilitating more reliable downstream analyses and cross-platform integration.

bioinformatics↗

bioRxiv2025

Functional rather than anatomic connectivity predicts seizure propagation in a multi-node model of focal neocortical epilepsy

Seizures propagate through the brain either locally or via widespread networks of anatomically and functionally connected nodes. These sites can be manipulated in the surgical treatment of human patients through ablation or stimulation. However, we still lack a full understanding of how seizures, ablation, and stimulation recruit or alter recruitment of these distant sites. Here, we apply widefield calcium imaging in a non-anesthetized rodent multi-nodal bilateral neocortical network model of focal epilepsy to examine excitatory and inhibitory cell recruitment. When we initiate seizures in somatosensory cortex (S1), they preferentially spread to an ipsilateral node in frontal cortex (M2) rather than across the corpus callosum to contralateral mirror somatosensory cortex. On the other hand, seizures rapidly spread across the corpus callosum in regions that connect M2 with its mirror M2 focus, indicating that this frontal region acts an amplifier for secondary generalization. Accordingly, ablation of M2 radically altered seizure propagation. Electrical stimulation of S1 revealed that S1 preferentially recruits excitatory cells in ipsilateral M2 but inhibitory cells in contralateral S1, which may explain the preferred propagation pathway. We also observed that the stimulation frequency can differentially determine the response of excitatory versus inhibitory neurons. Altogether, our findings suggest that seizures do not propagate homogeneously through anatomically connected nodes but are preferentially "pulled" to specific locations by excitatory/inhibitory balance. Thus, functional connectivity rather than anatomic connectivity will be more predictive of ictal spread, and more informative for ablative and stimulation-based therapeutics. SignificanceUnderstanding how seizures spread outward from one brain region to another is critical to informing new therapies for epilepsy patients. We used widefield calcium imaging and electrical stimulation to examine seizure propagation patterns in a bilateral brain network in non-anesthetized animals. We found that seizures unexpectedly preferred certain pathways over others, despite strong anatomical connections, and that these preferences were dictated by the type of cell (excitatory or inhibitory) recruited in distant nodes. We were able to reproduce these electrical imbalances with focal stimulation at varying frequencies. Our findings highlight that anatomy by itself is insufficient to fully identify seizure networks and suggests that specific neurostimulation parameters will confer different effects on distant brain regions. Key PointsO_LIFunctional connectivity rather than anatomic connectivity determines ictal propagation C_LIO_LIContralateral seizure propagation from S1 favors cross callosal spread in via motor, rather than somatosensory pathways. C_LIO_LIHigh frequency electrical stimulation differentially affects excitatory and inhibitory recruitment across a neocortical network C_LIO_LILow frequency electrical stimulation emulates seizure-like activity propagation across a network C_LIO_LIConnectome data may be insufficient to identify critical nodes and pathways of a seizure network C_LI

neuroscience↗

bioRxiv2024

Global Neuron Shape Reasoning with Point Affinity Transformers

Connectomics is a field of neuroscience that maps the brains intricate wiring diagram. Accurate neuron segmentation from microscopy volumes is essential for automating connectome reconstruction. However, state-of-the-art algorithms use image-based convolutional neural networks limited to local neuron shape context. Thus, we introduce a new framework that reasons over global neuron shape with a novel point affinity transformer. Our framework embeds a (multi-)neuron point cloud into a fixed-length feature set from which we can decode any point pair affinities, enabling clustering neuron point clouds for automatic proofreading. We also show that the learned feature set can easily be mapped to a contrastive embedding space that enables neuron type classification using a simple classifier. Our approach excels in two demanding connectomics tasks: correcting segmentation errors and classifying neuron types. Evaluated on three benchmark datasets derived from state-of-the-art connectomes, our method outperforms point transformers, graph neural networks, and unsupervised clustering baselines.

neuroscience↗

bioRxiv2021

Antibody Cocktail Exhibits Broad Neutralization against SARS-CoV-2 and SARS-CoV-2 variants

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has precipitated multiple variants resistant to therapeutic antibodies. In this study, 12 high-affinity antibodies were generated from convalescent donors in early outbreaks using immune antibody phage display libraries. Of them, two RBD-binding antibodies (F61 and H121) showed high affinity neutralization against SARS-CoV-2, whereas three S2-target antibodies failed to neutralize SARS-CoV-2. Following structure analysis, F61 identified a linear epitope located in residues G446 -S494, which overlapped with angiotensin-converting enzyme 2 (ACE2) binding sites, while H121 recognized a conformational epitope located on the side face of RBD, outside from ACE2 binding domain. Hence the cocktail of the two antibodies achieved better performance of neutralization to SARS-CoV-2. Importantly, F61 and H121 exhibited efficient neutralizing activity against variants B.1.1.7 and B.1.351, those showed immune escape. Efficient neutralization of F61 and H121 against multiple mutations within RBD revealed a broad neutralizing activity against SARS-CoV-2 variants, which mitigated the risk of viral escape. Our findings defined the basis of therapeutic cocktails of F61 and H121 with broad neutralization and delivered a guideline for the current and future vaccine design, therapeutic antibody development, and antigen diagnosis of SARS-CoV-2 and its novel variants.

immunology↗

bioRxiv2021

Ezh2 mediates epigenetic regulation of osteoclastogenesis and bone remodeling in mice

Osteoclasts derived from hematopoietic stem cells control bone resorption. Identifying novel molecules that can epigenetically regulate osteoclastogenesis has been an important basic and clinical issue. The polycomb group (PcG) protein enhancer of zeste homologue 2 (Ezh2), a histone lysine methyltransferase is associated with epigenetic regulation of numerous cellular processes, it is not yet clear on its involvement in bone cell development and homeostasis. Here, we crossed LysM-Cre mice with Ezh2flox/flox mice to delete Ezh2 in myeloid cell lineage mature macrophages. Conditional deletion of Ezh2 in macrophages resulted in significant increases in postnatal bone growth in the first 6 months of life, but tibia length and body weight gains were not different in knockout mice compared with their wild type controls. Significantly decreased osteoclastogenesis but increased bone mass without osteopetrosis were found in Ezh2 conditional knockout (CKO) mice. In contrast to female mice, one floxed Ezh2 gene copy recombinant with LysM-Cre+ (Ezh2flox/+LysM-Cre+) produced increased bone mass in young adult male mice compared with control mice (Ezh2flox/flox, LysM-Cre+ and wild type). Inflammatory milieu in bone was significantly lower in both male and female CKO mice compared with their respective controls. Deletion of Ezh2 in macrophages triggered increased gene expression of osteoclast suppressors, IRF8, MafB and Arg1 due to decreased Ezh2-induced trimethylation of H3K27me3. Conversely, NFATc1 and Cathepsin k expression were decreased. These findings suggest that pre-osteoclastic cell differentiation is under epigenetic control of osteoclast suppressive gene expression via an Ezh2-dependent mechanisms.

developmental biology↗

bioRxiv2020

Ultra-deep sequencing differentiates patterns of skin clonogenic mutations associated with sun-exposure status

Non-melanoma skin cancer is the most common human malignancy and is primarily caused by exposure to ultraviolet (UV) radiation. The earliest detectable precursor of UV-mediated skin cancer is the growth of cell groups harboring clonal mutation (CM) in clinically normal appearing skin. Systematic evaluation of CMs is crucial to understand early photo-carcinogenesis. Previous studies confirmed the presence of CMs in sun-exposed skin. However, the relationship between UV-exposure and the accumulation of CMs, and the correlation of CMs with skin cancer risk remain poorly understood. To elucidate the exact molecular and clinical effects of long-term UV-exposure on skin, we performed targeted ultra-deep sequencing in 450 individual-matched sun-exposed (SE) and non-sun-exposed (NE) epidermal punch biopsies obtained from clinically normal skin from 13 donors. A total of 638 CMs were identified, including 298 UV-signature mutations (USMs). The numbers of USMs per sample were three times higher in the SE samples and were associated with significantly higher variant allele frequencies (VAFs), compared with the NE samples. We identified genomic regions in TP53, NOTCH1 and GRM3 where mutation burden was significantly associated with UV-exposure. Six mutations were almost exclusively present in SE epidermis and accounted for 42% of the overall difference between SE and NE mutation burden. We defined Cumulative Relative Clonal Area (CRCA), a single metric of UV-damage calculated by the overall relative percentage of the sampled skin area affected by CMs. The CRCA was dramatically elevated by a median of 11.2 fold in SE compared to NE samples. In an extended cohort of SE normal skin samples from patients with a high- or low-burden of cutaneous squamous cell carcinoma (cSCC), the SE samples in high-cSCC patients contained significantly more USMs than SE samples in low-cSCC patients, with the difference mostly conferred by mutations from low-frequency clones (defined by VAF≤1%) but not expanded clones (VAF>1%). Our studies of differential mutational features in normal skin between paired SE/NE body sites and high/low-cSCC patients provide novel insights into the carcinogenic effect of UV exposure, and suggest CMs might be used to develop novel biomarkers for predicting cancer risk.Significance statement In UV radiation exposed skin, mutations fuel clonal cell growth. We established a sequencing-based method to objectively assess the mutational differences between sun-exposed (SE) and non-sun-exposed (NE) areas of normal human skin. Striking differences, in both the numbers of mutations and variant allele frequencies, were found between SE and NE areas. Furthermore, we identified specific genomic regions where mutation burden is significantly associated with UV-exposure status. These findings revealed previously unknown mutational patterns associated with UV-exposure, providing important insights into UV radiation’s early carcinogenic effects. Additionally, in an extended cohort, we identified preliminary association between normal skin mutation burden and cancer risk. These findings pave the road for future development of quantitative measurement of subclinical UV damage and skin cancer risk.Competing Interest StatementThe authors have declared no competing interest.Common AbbreviationsUVUltravioletCMClonogenic mutationNMSCNonmelanoma skin cancerSESun-exposedNENon-sun-exposedUSMUV-signature mutationNUSMNon-UV-signature mutationCRCACumulative Relative Clonal AreacSCCCutaneous squamous cell carcinomaAKActinic keratosisSNVSingle nucleotide variant Indels – Insertions/deletionsDNVDinucleotide variantCSNVCluster of single nucleotide variantMACMulti-Nucleotide Variant Annotation CorrectorVAFVariant allele frequencyView Full Text

cancer biology↗