Biology subjects

Zhou, Z.

Publications and source records attributed to Zhou, Z..

At least 55 records · Page 3Linked to original sources

bioRxiv2026

Restoring hippocampal sharp-wave ripples under glutamate transporter dysfunction

Disruption of glutamate homeostasis is believed to contribute to the early progression of Alzheimers disease (AD) and associated neurodegeneration. Soluble amyloid-{beta} oligomers impair excitatory amino acid transporters (EAATs), reducing glutamate clearance, while also enhancing glutamate release from neurons and astrocytes. Together, these effects produce persistent glutamatergic dysregulation that disrupts synaptic and network function. Here, we asked whether the effects of EAAT attenuation can be mitigated through ion-channel modulation. TBOA, a selective EAAT inhibitor, was used to model early-stage glutamatergic dysregulation. TBOA reduced the local field potential amplitude of hippocampal sharp-wave ripples (SWRs) in mouse hippocampal slices, suggesting that glutamate accumulation disrupts network synchrony. Calcium imaging further showed that TBOA diminished SWR-associated population calcium transients while promoting spontaneous calcium transients in individual neurons, indicating a shift from coordinated population activity toward disorganized cellular activity. KCNQ-channel openers ML213 and ICA-27243 partially restored the TBOA-induced decline in SWR amplitude. In contrast, similar restorative effects were not observed following modulation of other ion channels, including blockade of AMPA and NMDA receptors or HCN/Ih channels, or activation of large-conductance Ca2+-activated K+ (BK) channels and G-protein-activated inwardly rectifying potassium (GIRK) channels. Together, these findings suggest that KCNQ-channel openers may occupy a unique position in mitigating glutamate-related hyperexcitability during early AD-associated network dysfunction.

neuroscience↗

bioRxiv2026

Faithful Supervised Dimensionality Reduction for Biomedical Data via Decision Geometry

Unsupervised dimensionality reduction methods aim to preserve intrinsic data geometry by maintaining local neighborhoods and approximate global relationships in low-dimensional embeddings, but they do not use label information and therefore may fail to reflect task-relevant class structure in biomedical and health applications. Supervised dimensionality reduction (SDR) incorporates labels to improve class organization, yet existing approaches often face a trade-off between discrimination and geometric faithfulness. Linear supervised methods are stable and interpretable but are limited in their ability to capture nonlinear structure, whereas many nonlinear methods impose supervision directly in the embedding space, which can over-separate classes and distort the underlying manifold. In biomedical applications, labels such as cell types in single-cell data or patient status in clinical cohorts provide meaningful biological signal, and supervised dimensionality reduction can use this information to produce more informative low-dimensional representations. Here we propose a new framework, DG-UMAP (Decision-Geometry UMAP), for faithful supervised dimensionality reduction via decision geometry. We first fit a classifier in the original feature space and use its boundary-local decision geometry to construct a low-rank metric deformation that emphasizes discriminative directions while limiting geometric distortion. Parametric UMAP is then applied to the transformed space, so supervision acts through the ambient geometry rather than by directly forcing class separation in the embedding. Across synthetic and multiple real-world biomedical datasets, our method yields embeddings with improved agreement with class structure and global organization while preserving local neighborhood quality.

bioinformatics↗

bioRxiv2026

Spatially resolved transcriptomic identification of thousands of neurons recorded in vivo.

Transcriptomics has transformed our understanding of the brain, but assigning transcriptomic identities to neurons recorded in vivo remains challenging at scale. Existing platforms can pair transcriptomic identity with two-photon calcium imaging in small populations of approximately 100 neurons, but they require recorded cells to be sparse and therefore cannot be applied to large population recordings. Here, we present coppaFISH 3D, a spatially resolved transcriptomics method, and CASTalign, an in silico alignment framework, which together enable transcriptomic identification of thousands of simultaneously recorded cells. coppaFISH 3D detects hundreds of genes in thick 50m fixed sections while preserving tissue integrity, enabling both 3D registration to in vivo imaging and integration with immunofluorescence labelling. The platform is fully powered by open chemistry and open source software, runs on commodity hardware, and can be performed at very low cost per section. It therefore enables transcriptomic identification of recorded neurons at scale, making it possible to study how transcriptomic identity shapes activity in neural populations.

neuroscience↗

bioRxiv2026

Endogenous network modeling reveals mechanisms of repair Schwann cell decline and potential recovery targets

Schwann cells, the principal glial cells of the peripheral nervous system, play a central role in nerve repair following injury. Upon injury, mature Schwann cells dedifferentiate into repair Schwann cells. These processes are governed by complex gene regulatory networks, yet the quantitative dynamics of these processes remain unclear. Here, using a bottom-up systems biology approach, we constructed an endogenous regulatory network model based on experimentally validated interactions, without relying on high-throughput data as input. The model captures Schwann cell dedifferentiation dynamics and reveals a potential landscape composed of stable states and intermediate transition states. Simulations recapitulate post-injury trajectories and confirm the role of c-Jun upregulation in maintaining repair capacity. Furthermore, the model predicts multiple potential therapeutic targets, including P53, JNK, and PTEN, for sustaining repair competence. We also identify intrinsic heterogeneity within repair Schwann cells and uncover key transition states that simultaneously connect repair-competent cells to both repair-deficient and apoptotic phenotypes, indicating that these intermediate states may represent critical regulatory bottlenecks and key cellular targets for improving the success of peripheral nerve regeneration. Overall, this work provides new insights into the precise regulation of Schwann cell fate and establishes a theoretical framework for regenerative medicine and clinical strategies in peripheral nerve repair.

neuroscience↗

bioRxiv2026

A Decade of Deep Learning-based Biomedical Image Segmentation

Biomedical image segmentation is a fundamental problem in computational biomedicine that aims to precisely delineate anatomical and biological structures, tissue types, or pathological regions in biomedical images. Accurate segmentation is essential for interpretation, decision-making, and quantitative analysis across a wide range of biological and medical applications. Over the past decade, the field has undergone a profound paradigm shift, evolving from task-specific specialist models to universal foundation models. This review provides an in-depth analysis of the evolution, tracing how the limitations of local discriminative learning drove the transition toward transformer-based global modeling, and large-scale generative pre-training. To help navigate the diverse landscape of interaction paradigms, we introduce the first systematic taxonomy of promptable biomedical image segmentation, categorizing existing methods into six distinct types, enabling users to intuitively select appropriate prompting strategies based on visual demonstrations and quickly pinpoint relevant literature (Prompt Type Visualization). Beyond model architectures, we discuss parallel advancements in dataset development, evaluation protocols, and application-specific adaptations across radiology, pathology, and biology. Integrating these powerful foundation models with rigorous domain-specific adaptation has great potential to improve patient outcomes and healthcare efficiency. Finally, we highlight key challenges in trustworthiness and clinical integration that must be overcome to realize the potential of the next generation of biological and medical generalists.

bioengineering↗

bioRxiv2026

Spatial Multi-Omics Reveals Extracellular Matrix Remodeling and VSMC Phenotypic Switching in Moyamoya Disease

BackgroundMoyamoya disease (MMD) is a progressive cerebrovascular disorder characterized by steno-occlusive lesions and intimal hyperplasia. Although vascular smooth muscle cell (VSMC) phenotypic switching is implicated in its pathogenesis, the precise spatial interplay between extracellular matrix (ECM) remodeling and local metabolic alterations within the distinct vascular microenvironments remains unknown. MethodsSuperficial temporal artery (STA) samples from patients with MMD and controls were analyzed by histology, immunofluorescence, spatial transcriptomics, spatial proteomics, and spatial metabolomics. Single cell RNA sequencing was used to profile the cellular landscape of STA tissues. To functionally validate the identified pathway, human brain vascular smooth muscle cells (HBVSMCs) were stimulated with fibronectin 1 (FN1), and patient derived induced pluripotent stem cell smooth muscle cells (iPSC-SMCs) were generated for migration and protein expression assays following ITGA5 silencing or focal adhesion kinase (FAK) inhibition. ResultsMMD STA samples exhibited marked intimal hyperplasia with medial thinning and intimal accumulation of -SMA positive cells. Spatial transcriptomic and proteomic analyses identified an intimal remodeling program characterized by increased FN1, EFEMP1, fibronectin, ITGA5, and FAK, together with reduced MYH11. FN1 stimulation promoted smooth muscle cell migration, ECM associated protein expression, and FAK phosphorylation, whereas ITGA5 knockdown or FAK inhibition attenuated these effects. Patient derived MMD iPSC-SMCs showed similar abnormalities, including enhanced migration, increased FAK activation, reduced contractile markers, and increased ECM associated proteins. Spatial metabolomics and integrated multi-omics analyses further revealed that these changes were coupled to a metabolically depleted intimal niche enriched for reduced acyl-CoA related metabolites. ConclusionsIntegrated spatial multi-omics identifies coupled ECM remodeling and metabolic alteration in the hyperplastic intima of MMD. Within this context, the FN1-ITGA5-FAK axis emerges as a plausible mediator of smooth muscle remodeling that warrants further validation. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=165 SRC="FIGDIR/small/721225v1_ufig1.gif" ALT="Figure 1"> View larger version (90K): org.highwire.dtl.DTLVardef@3f6e19org.highwire.dtl.DTLVardef@554d49org.highwire.dtl.DTLVardef@451dd9org.highwire.dtl.DTLVardef@1aac1e9_HPS_FORMAT_FIGEXP M_FIG C_FIG

neuroscience↗

bioRxiv2026

Epigenetic and 3D Genome Changes Drive Primary Trastuzumab Resistance in HER2+ Breast Cancer

Primary trastuzumab resistance remains a major challenge in the treatment of HER2-positive breast cancer, with limited therapeutic options highlighting the need for deeper mechanistic insight. Here, integrative analysis of epigenomic and three-dimensional chromatin architecture reveals widespread remodeling of the regulatory landscape in resistant cells. Alterations in promoter-associated histone modifications?particularly H3K4me3 and H3K27me3?together with changes in active enhancer activity and promoter-enhancer interactions, appear to be key drivers of transcriptional reprogramming underlying resistance. As an illustrative example, SGK1 exhibits epigenetic activation in resistant cells, characterized by increased promoter H3K4me3 and enhanced chromatin interactions, consistent with its role in supporting cell survival, proliferation and metastasis. Overall, our findings highlight epigenetic and chromatin structural alterations as central mechanisms of primary trastuzumab resistance, providing a framework for novel therapeutic strategies development.

cancer biology↗

bioRxiv2026

A Noncontiguous Code for RNA-Guided DNA Recognition Preceded CRISPR

CRISPR-Cas systems use RNA-guided proteins for adaptive immunity through a mechanism whose origin is unknown. Here we report the discovery of Viral Interference Programmable Repeat (VIPR) systems consisting of a Vipr protein more ancient than CRISPR-Cas and vrRNAs comprising alternating GGY/NN motifs. Unlike canonical guide RNAs that base pair with nucleic acid targets using an uninterrupted sequence, vrRNAs recognize double-stranded DNA through a noncontiguous code in which the variable NNs of each repeat collectively specify a target that itself contains a gapped recognition sequence. Analysis of natural vrRNA targets suggests VIPR acts against competing phages. We demonstrate programmable phage defense by redirecting the complex for transcriptional repression. These results suggest that the roots of adaptive immunity lie in ancient warfare between viruses, and reveal a new logic for programmable genetic control.

molecular biology↗

bioRxiv2026

Connections across regional glymphatic clearance, neural activity and amyloid-β deposition in cortex

Neural activity inevitably produces waste, which promotes neurodegeneration with topographic features. The glymphatic system is important for waste clearance. However, the spatial characteristics of glymphatic clearance across cortex and whether it interplays with neural activity in contribution to amyloidosis in human remain unexplored. Here, by intrathecal administration of gadolinium-based contrast agents, glymphatic influx and clearance patterns across cortex in 96 participants are depicted via Glymphatic MRI. Analyses integrating post-mortem transcriptomic profiles from Allen Human Brain Atlas indicate that, genes related with excitatory and inhibitory neurons, and pathways engaging in synaptic function were enriched in regions with faster glymphatic clearance. FALFF was calculated from resting-state fMRI to represent neural activity. At the regional level, based on a subgroup with rs-fMRI (N = 15), regional glymphatic clearance was positively coupled with spontaneous neural activity. Mismatch index, reflecting decoupling between spontaneous neural activity and glymphatic clearance function, turned out to be positively associated with regional severity of amyloidosis using open-source 11C-PiB dataset. Together, this study for the first time demonstrates the intricate interplays between neural activity and glymphatic dynamics from transcriptional to physiological level. The mismatch between these two processes may serve as an undescribed comprehensive mechanism promoting regional vulnerability to proteopathy and subsequent neurodegeneration in cortex.

neuroscience↗

bioRxiv2026

CHORD: a framework for cross-species single-cell integration across gene, cell and cell-type levels

Quantifying cross-species relationships among cell types from single-cell transcriptomic data can reveal both conserved and divergent patterns of cell-type hierarchies. However, existing cross-species integration methods can be limited in modeling genes beyond orthologs by leveraging cell-type-resolved transcriptional context, or in learning explicit type-level representations. Here we present CHORD, a cross-species integration framework that jointly learns representations of genes, cells and cell types. We demonstrate that CHORD can integrate cross-species single-cell atlases and support cell-type annotation with unknown cell-type detection. In the frog-zebrafish embryogenesis and mammalian motor cortex atlases, CHORD infers cell-type trees that place conserved cell types from different species in relative proximity and summarize hierarchical relationships among cell types. CHORD also supports cross-species comparison of continuous phenotypic variation by placing embryonic cells along an aligned developmental timeline. CHORD further yields gene embeddings that capture orthologous and functional relationships, and gene importance scores linking genes to cell types.

bioinformatics↗

bioRxiv2026

Hybrid Gated Fusion: A Multimodal Deep Learning Framework for Protein Function Annotation

Protein function annotation requires integrating diverse biological signals, yet existing multimodal methods often struggle with missing inputs and redundant information. We present Hybrid Gated Fusion, a multimodal architecture that combines intrinsic protein features, including sequence and structure, with extrinsic functional context from text and interaction networks. Rather than weighting all modalities equally, the model uses bilinear gating to assess both the informativeness of each modality and its agreement with the others, while auxiliary supervision reduces modality dominance and preserves useful signal in weaker modalities. On the CAFA3 benchmark, a single Hybrid Gated Fusion model achieves state-of-the-art performance in Biological Process (Fmax = 0.601) and Cellular Component (Fmax = 0.706), while remaining competitive in Molecular Function (Fmax = 0.702). Analysis of the learned gates shows that interaction networks and text often provide complementary functional signals, whereas structural features are down-weighted when redundant but remain valuable under sparse-input settings. These results establish Hybrid Gated Fusion as a robust and scalable framework for genome-scale protein function annotation. Availability and implementationSource code and reproduction scripts are freely available at https://github.com/psipred/PFP. Pre-computed embeddings, data splits, and model checkpoints are deposited at https://doi.org/10.5281/zenodo.19498341.

bioinformatics↗

bioRxiv2026

The influence of tension-compression switches on brain anisotropic modelling

Finite element (FE) head models are valuable tools for investigating brain injury mechanics, with their reliability critically dependent on accurate material modelling. White matter (WM) is often considered mechanically anisotropic due to its aligned axonal fiber architecture and is commonly represented using fiber-reinforced hyperelastic formulations such as the Gasser-Ogden-Holzapfel (GOH) model. A fundamental assumption of the GOH model is that fibers contribute only in tension and not in compression, requiring the use of tension-compression switches. However, inconsistencies were noted in the formulation of tension-compression switches with the influence on computational biomechanics unknown. To address this knowledge gap, three commonly used switching schemes - differing in both the switching parameter and the treatment of compressed fibers - were theoretically elaborated and numerical implementation within the GOH framework to simulate the mechanical anisotropy of WM in impact simulations. Results from the case-based and group-level analyses demonstrated that both the switching parameter and the treatment of compressed fibers affected WM deformation. Significant cross-scheme strain differences were noted in the first principal strain at the element level and fiber strain at the fiber level. These findings highlighted the mechanical role of tension-compression switch in the GOH-based brain modelling and advocated the adoption of fiber stretch itself as the switching parameter to discriminate the tensile and compressive fibers. The current study provides important guidance for the anisotropic constitutive models in brain tissue and calls for direct verification of the tension-compression switch hypothesis in axonal fibers.

biophysics↗

bioRxiv2026

Rapid Remodeling of Human White Adipose Tissue Following Bariatric Surgery

Bariatric surgery induces profound weight loss and improvement of obesity-associated metabolic dysfunction. Recent studies have shown that adipose tissue undergoes remodeling after weight loss, characterized by a reduction in proinflammatory immune cells, increased vascularization, and a shift in the adipocyte transcriptome, but these studies focused on time points long after surgery. We performed single nucleus RNA-seq (snRNA-seq) in subcutaneous white adipose tissue (SAT) samples from subjects with obesity undergoing bariatric surgery, collected at baseline and at one, six, and twelve months after surgery. We identify profound remodeling of SAT within the first month after surgery, characterized by a surge in lipid-associated macrophages and sharp reductions in specific populations of adipocytes, adipose stromal and progenitor cells (ASPCs), and endothelial cells. Transcriptional profiles strongly suggest that some adipocytes undergo apoptosis soon after surgery, while new adipocytes are generated by de novo differentiation. Mechanistically, the data are consistent with a model whereby coordinated early loss of a hedgehog signaling axis between endothelial cells and an anti-adipogenic population of ASPCs known as adipose regulatory cells (Aregs; ASPCPTCH2) enables a transient burst of adipogenesis to occur. Interestingly, very few of these early features seen in human subjects are represented in a mouse model of surgical weight loss.

physiology↗

bioRxiv2026

TopicVI: A Knowledge-guided deep interpretable model for resolving context-specific gene programs

Mechanistic insights from single-cell and spatial transcriptomics largely rely on cell clustering, differential expression analysis, and interpretation through prior biological knowledge. However, this approach is often limited by the reliance on curated biological priors that fail to capture context-specific gene programs, particularly in complex disease states. To address this gap, we introduce TopicVI, a deep interpretable model that integrates established biological knowledge with data-driven refinement to discover context-dependent gene programs in single-cell and spatial transcriptomic data. TopicVI jointly infers cell clusters and gene topics using optimal transport to flexibly align prior gene programs with observed data while permitting context-specific refinements. Comprehensive benchmarking demonstrates that TopicVI outperforms existing methods in biological conservation, batch correction, topic coherence, and rare cell identification. TopicVI effectively disentangles multiple sources of biological variation, such as separating anatomy-specific expression patterns from disease-associated signatures in spatial transcriptomics. Applying TopicVI to glioblastoma datasets, we identify gene topics related to cell cycle regulation and EGFR signaling that reveal convergent tumor states across distinct drug perturbations. By integrating prior knowledge with data-driven discovery, TopicVI enables identification of interpretable gene programs that illuminate biological processes and therapeutic mechanisms in complex transcriptomics data.

bioinformatics↗

bioRxiv2026

GYDE: A collaborative drug discovery platform for AI-powered protein design and engineering

As computational tools and machine learning models for protein sciences continue to advance and proliferate, bench scientists face increasing technical challenges adopting these tools for specific applications such as drug discovery. Here we present GYDE (Guide Your Design and Engineering), an open-source, versatile, and web-based collaboration platform designed to make computational analyses of proteins and antibodies easily accessible to bench scientists. GYDE enables the exploration of sequence-structure-function relationships through a tightly integrated visual interface, offering researchers a comprehensive exploration of protein functional determinants either via real assay data or computational tools. GYDEs intuitive interface facilitates seamless access to cutting-edge AI models for protein and antibody structure prediction, design, and downstream analyses. The flexible and easy addition of new tools and models is facilitated by the use of the Slivka compute API. The platform supports saved sessions that enable researchers to easily share their findings with other users, fostering a more collaborative scientific community. GYDE is freely available for protein scientists in academia and industry to build drug discovery analytics platforms customized to their needs.

bioinformatics↗

bioRxiv2026

Metabolic specialization structures gut bacterial niches and drives colorectal cancer progression

Despite the established association between the gut microbiome and colorectal cancer (CRC), the functional distinction between microbial passengers and drivers of CRC progression remains unresolved. Here, we collected stool, blood, as well as paired tumor, and normal mucosa tissues from seventy-seven CRC patients to characterize the systemic and localized impact of the gut microbiome on early- and late-stage CRC. By deep shotgun metagenomic sequencing, we identified distinct bacterial species and functions residing in tumor versus normal mucosa, highlighting an enrichment of oral-associated bacteria in tumor tissues. Several of these species remained undetected in the stool microbiome analysis. We further combined bacterial culturing with untargeted metabolomics of bacteria enriched in tumor and normal mucosa tissues, revealing distinct clusters of metabolic potential. Functional testing of multiple members from one cluster comprising both tumor- and mucosa-enriched species revealed Leptotrichia wadei as a pro-tumorigenic bacterium in a murine CRC model. Single-nucleus RNA sequencing and in vitro experiments further demonstrated that L. wadei and its secretome induces M2 macrophage polarization to promote tumor growth. Overall, our study shows that metabolic specialization structures microbial colonization niches, while species-specific metabolic outputs identify functional drivers of CRC progression, and uncovers L. wadei as an oncogenic bacterium in CRC.

microbiology↗

bioRxiv2026

BAF complexes maintain accessibility at stimulus-responsive chromatin and are required for transcriptional stimulus responses

BackgroundGene expression changes in response to developmental and environmental cues rely on cis-regulatory sequence elements (cREs). BRG1/BRM-Associated Factors (BAF) chromatin remodeling complexes maintain chromatin accessibility at many cREs, enabling binding by transcription factors (TFs). However, cREs exhibit a broad range of sensitivity to loss of BAF function, and the basis of this variability remains unknown. ResultsTo identify the characteristics of BAF-dependent cREs, we mapped chromatin accessibility changes following acute pharmacologic BAF inhibition in GM12878 lymphoblastoid cells. We integrated these results with over 100 TF and histone modification ChIP-seq datasets and used machine learning to identify features that predict chromatin accessibility changes. We found that Activator Protein 1 (AP-1) factors and lymphoid lineage-defining TFs including RUNX3 and PU.1 predicted BAF-dependence. Strikingly, we found that cREs bearing the chromatin signature of "primed" enhancers - enriched for H3K4me1 but lacking H3K27ac - were significantly more sensitive to BAF inhibition than typical active enhancers. As primed enhancers are known to facilitate transcriptional responses to stimuli, we tested the requirement of BAF activity in these responses. Acute BAF inhibition was sufficient to prevent both chromatin and transcriptional responses to interferon gamma and dexamethasone. cREs which normally gained accessibility in response to stimulation failed to do so with BAF inhibition, and these cREs were linked to genes with suppressed transcriptional induction. ConclusionsCollectively, our results demonstrate a requirement for continuous BAF activity to enable stimulus response and suggest that defective signal responsiveness may be a pathogenic mechanism in disease states caused by loss-of-function mutations in BAF subunits.

genomics↗

bioRxiv2026

HViLM: A Foundation Model for Viral Genomics Enables Multi-Task Prediction of Pathogenicity, Transmissibility, and Host Tropism

MotivationThe emergence of novel viral pathogens poses critical threats to global health, yet current computational approaches for viral risk assessment are predominantly virus-specific and require extensive retraining for each new threat. Computational methods for rapid characterization of emerging viruses across multiple epidemiologically relevant dimensions--pathogenicity, host tropism, and transmissibility--are urgently needed to inform public health responses and guide experimental prioritization. ResultsWe present HViLM (Human Virome Language Model), the first foundation model for panviral genomic analysis through continued pre-training of DNABERT-2 on 5 million non-redundant viral sequences (MMseqs2-clustered from 25 million chunks at 80% identity) spanning 9,000 species across 45+ viral families from the VIRION database. We introduce the Human Virome Understanding Evaluation (HVUE) benchmark comprising seven curated datasets across three prediction tasks: pathogenicity classification, host tropism prediction, and transmissibility assessment. Through parameter-efficient fine-tuning with LoRA, HViLM achieves state-of-the-art performance with average accuracies of 95.32% for pathogenicity, 96.25% for host tropism, and 97.36% for transmissibility assessment. The model demonstrates robust cross-family generalization, substantially outperforming sequence-similarity baselines and general genomic foundation models. Attention-based interpretability analysis reveals that HViLM captures biologically meaningful pathogenicity determinants through molecular mimicry of host regulatory elements, including convergent evolution of eight independent sequences targeting Interferon Regulatory Factor 1 (Irf1) for immune evasion. AvailabilityThe HVUE benchmark datasets, training scripts, and complete implementation are publicly available at https://github.com/duttaprat/HViLM. Pre-trained HViLM-base model weights and fine-tuned task-specific variants are available on Hugging Face at https://huggingface.co/duttaprat/HViLM-base. Supplementary informationSupplementary data are available online.

bioinformatics↗