Biology subjects

Smith, K.

Publications and source records attributed to Smith, K..

At least 37 records · Page 2Linked to original sources

bioRxiv2025

Human C. difficile-specific memory B cells encode protective IgG1 despite predominance of non-neutralizing antibodies.

Clostridioides difficile remains a common source of nosocomial infection resulting in a wide range of clinical outcomes and for which there is no vaccine and limited therapeutic options. C. difficile Toxin B (TcdB)-specific IgG is the best correlate of protection against severe and recurrent disease. However, there are very few therapeutic IgG fully human monoclonal antibodies (hmAbs) that have shown efficacy thus far. We therefore hypothesized that the memory B cell (Bmem) compartment of individuals who have recovered from infection may be dominated by non-protective IgG molecules but encode some antibodies that are protective. We therefore produced hmAbs from a library of Bmem-encoded IgG1 sequences. Most TcdB-specific hmAbs displayed low affinity binding and poor neutralization of TcdB in vitro while a few hmAbs had high affinity binding and good TcdB neutralization. The results correlated with in vivo studies in which high affinity, TcdB-neutralizing hmAbs provided moderate protection of C57Bl/6 mice against C. difficile disease. Similar protection was observed in Tg32 mice expressing the human FcRn transgene, indicating that gut delivery of hmAb did not account for limited efficacy. Although non-protective antibody sequences dominate the repertoire, human Bmem cells may be a good source of therapeutic hmAbs for C. difficile treatment.

immunology↗

bioRxiv2025

Human stem cell models for group 3 medulloblastoma uncover JARID1B as a regulator of the chromatin landscape.

Medulloblastoma (MB) is one of the most prevalent malignant brain tumors in children, with tremendous cognitive and neuroendocrine disability among survivors. Group 3 MB (G3MB) has poor overall survival at <50%, high frequencies of metastases, and no targeted therapies. Amplification of MYC and activation of TGF{beta} signaling occur frequently in G3MB. Many tumors have no reported mutations, suggesting epigenetic drivers. We here describe novel humanized models for G3MB from human induced pluripotent stem cells (hiPSC). By transducing hiPSC-derived neuroepithelial stem cells (NESC), we determined that: 1) both MYC and TGF{beta} effectors drove tumors in vivo; 2) MYC/TGF{beta}R1 in combination led to more aggressive tumors and resistance to clinical inhibitors of TGF{beta}, and 3) NESC-derived tumors clustered with human G3MB. To decipher mechanisms, we integrated RNA-sequencing with CUT&RUN (for MYC genomic localization and post-translational modification of histones). MYC-bound neural developmental genes were repressed in MYC/TGF{beta}R1 co-driven lines. Gene signatures associated with the Polycomb Repressive Complex (PRC) demarcated with H3K27me3; the histone mark directly regulated by PRC. We identified JARID1B, a MYC binding partner and H3K4me3 demethylase, as a regulator of repressed neural genes. Primary G3MB also showed increased levels of H3K27me3 concurrent with higher expression of JARID1B. Knockdown of JARID1B in human G3MB cell lines reduced growth, supporting potential as a therapeutic target. We conclude that a MYC-TGF{beta}-JARID1B axis represses target genes to drive G3MB and present new humanized models for G3MB to understand epigenetic dysregulation in G3MB.

cancer biology↗

bioRxiv2025

Persistent CD4+ T cell functional deficits during recovery from prolonged symptomatic SARS-CoV-2 infection

Symptoms of acute SARS-CoV-2 infection often resolve quickly but are sometimes associated with persistent immune dysfunction. The factors that predispose individuals to compromised immune function have not been well defined. We investigated CD4+ T cell phenotype and function in a small cohort of individuals who recovered from mild to moderate SARS-CoV-2 infection without hospitalization and were divided into short or prolonged symptom duration groups. Five individuals with prolonged symptom duration showed marked downregulation of CD4 on CD3+CD8- T cells (CD4low group) and a poor response to TCR stimulation with the superantigen Staphylococcal enterotoxin B (SEB), as shown by weak upregulation of the activation markers CD134 and CD69. CD4 surface intensities recovered to normal levels in four of these individuals within 3-12 months. Selected cytokines (IL-1RA, IL-7, and VEGF) were elevated in individuals with low CD4, but plasma levels of anti-S1 IgG did not correlate with CD4 defects. Bulk RNA sequencing of unstimulated and SEB-treated CD3+CD8- T cells revealed a >50% reduction in the number of differentially expressed genes in the CD4low group compared to the same individuals after CD4 levels were recovered and a healthy control group. Upstream regulator analysis of differentially expressed genes in unstimulated CD4low cells suggested a response to IFN, while SEB-stimulated CD4low cells showed reduced functionality of IL-2, CD28, and SATB1 regulated pathways. In summary, prolonged symptomatic recovery from SARS-CoV-2 infection was associated with a global CD4+ T cell response defect, defined by low surface CD4 expression, evidence of IFN signaling, and defective T cell activation.

immunology↗

bioRxiv2025

Proteostasis Modelling using Deuterated Water Metabolic Labeling and Data-Independent Acquisition Mass Spectrometry

We present the first application of a deuterated water metabolic labeling workflow coupled with data-independent acquisition (DIA) tandem mass spectrometry (MS/MS) for quantifying label enrichment in MS/MS to study protein turnover. The approach automates the turnover rate determination from combined precursor and fragment ions. The truncation of the observed isotope distributions of fragments is overcome by implementing an approach to determining the label enrichment from two mass isotopomers. The high redundancy of fragment ions provides a confident assessment for deuterium enrichment. The DIA approach provides increased proteome coverage and depth for quantifying protein turnover rates compared to the traditional data-dependent (DDA) approach. This novel approach is validated in a murine myotube model of muscle hypertrophy and atrophy through the treatment with insulin-like growth factor 1 (IGF-1) and dexamethasone (Dex), respectively.

bioinformatics↗

bioRxiv2025

Locally persistent lineages are an important component of zoonotic foodborne pathogen Shiga toxin-producing Escherichia coli O157:H7 disease ecology

Evidence suggests that Shiga toxin-producing Escherichia coli (STEC) strains do not persist at the farm level. We hypothesized that ecosystem-level STEC persistence occurs and contributes significantly to disease burden. We tested this by identifying locally persistent lineages (LPLs) of STEC O157:H7 in Minnesota. We identified 15 distinct LPLs, which were associated with 35.3% of reported cases in Minnesota and persisted for 1.3 to 8.6 years. LPLs were associated with multiple outbreaks with Minnesota sources and no multi-state outbreaks, and LPL cases were spatially clustered. Our findings show long-term persistence in defined geographic areas, suggesting the importance of ecosystem-level persistence. SummaryShiga toxin-producing Escherichia coli (STEC) persistence is poorly understood. This study examined long-term STEC persistence. Locally persistent lineages (LPLs) of STEC O157:H7 contributed significantly to disease burden and persisted up to 8.6 years. LPLs suggest the importance of ecosystem-level persistence.

genomics↗

bioRxiv2025

Single-cell neural network classifiers reveal that PM21 NK cell expansion is dependent on B cell signaling

In the field drug development ML/AI methods are being applied to improve drug production speed, costs, and reliability. In allogenic NK cell therapy production, one of the biggest challenges is the inherent variability in the donors that provide the starting material for NK cell expansion. In this study we performed PM21-mediated NK cell expansion on 26 donors, and in parallel performed single-cell transcriptomics on the same donor sample prior to expansion. Canonical differential expression analysis and cell state abundance did not highlight any significant difference between donors with high and low NK cell expansion yield. Instead, training neural networks classifiers for high-yield donors enabled identifying several highly predictive models with perfect cross-validation recall. Further investigation of the most predictive models unveiled a previously unknown role for B cell in supportive NK cell expansion. Overall, this study represents a blueprint for combining deep phenotyping and machine learning methods to unveil novel biology and improve the quality and speed of delivery of cell therapeutics to patients.

systems biology↗

bioRxiv2025

Discovery of homoharringtonine pathway enzymes reveals a whole plant model for coordinated biosynthesis

Plants produce diverse molecules that inhibit protein translation. A lead example is homoharringtonine (HHT), a key tool for ribosomal profiling and an FDA-approved treatment for chronic myeloid leukemia. HHT is commercially produced through semi-synthesis from the alkaloid core cephalotaxine (CET) extracted from endangered Cephalotaxus species. Despite its significance, the CET/HHT biosynthetic pathway remains unresolved. Here, we use paired untargeted metabolomics (stable-isotope precursor feeding) and transcriptomics to elucidate a near-complete biosynthesis to CET without prior knowledge of intermediates and biosynthetic genes. We show that while the CET core is biosynthesized only in growing root tips, CET and HHT accumulate throughout the plant. We discovered seven pathway intermediates and six novel enzymes that produce cephalotaxinone, the likely direct precursor of CET. Included are non-canonical cytochrome P450s, an atypical short-chain dehydrogenase, and a 2-oxoglutarate-dependent dioxygenase that together result in carbon excision and CET/HHT pentacyclic backbone formation. This study establishes a metabolic route to the HHT core scaffold and suggests a whole-plant coordination model in Cephalotaxus, where cephalotaxinone is produced in root tips and distributed throughout the plant for subsequent elaboration to HHT.

biochemistry↗

bioRxiv2025

Dysferlin Regulates Cardiac T-tubule Structure and Excitation-contraction Coupling in Isolated Cardiac Myocytes at Rest and in Response to Acute Hypo-osmotic Stress and is Protective Against Arrhythmias in Langendorff-perfused Hearts

Dysferlin is a membrane-associated protein that supports skeletal muscle function such that mutations in the DYSF gene can cause muscular dystrophy. Growing evidence suggests dysferlin regulates cardiac function, but this is less well understood. Tight regulation of the cardiac transverse-(T)-tubule network and excitation-contraction (EC) coupling mechanism is essential for healthy cardiac physiology. Remodelling of T-tubules and the EC coupling mechanism is observed during periods of cardiac stress and pathologies, promoting arrhythmias. However, little is known about how these processes are regulated and any protective mechanisms which may limit detrimental effects. Using a global dysferlin knockout (KO) mouse we have shown that the loss of dysferlin leads to a decrease in T-tubule density and the amplitude and rate of decay of the systolic Ca2+ transient but a narrowing of the dyadic cleft. Electrical mapping of ex vivo DYSF KO hearts shows they are more susceptible to ventricular arrhythmias. To induce stress, we used hypo-osmotic shock injury (OSI) to damage T-tubule networks in cardiac myocytes, in vitro. OSI increased T-tubule fragmentation and caused dysregulation of intracellular Ca2+ handling in dysferlin KO cells relative to WT controls. Finally, we observed that a natural decline in WT cardiac dysferlin abundance, which may contribute to the natural age-dependent maladaptive T-tubule remodelling that occurs in the mammalian ventricle. In summary, these findings demonstrate an essential role for dysferlin in cardiac physiology, especially during conditions of stress, which is decreased in normal ageing.

physiology↗

bioRxiv2025

Rapamycin exerts its geroprotective effects in the ageing human immune system by enhancing resilience against DNA damage

mTOR inhibitors such as rapamycin are among the most robust life-extending interventions known, yet the mechanisms underlying their geroprotective effects in humans remain incompletely understood. At non-immunosuppressive doses, these drugs are senomorphic, i.e. they mitigate cellular senescence, but whether they protect genome stability itself has been unclear. Given that DNA damage is a major driver of immune ageing, and immune decline accelerates whole-organism ageing, we tested whether mTOR inhibition enhances genome stability. In human T cells exposed to acute genotoxic stress, we found that rapamycin and other mTOR inhibitors suppressed senescence not by slowing protein synthesis, halting cell division, or stimulating autophagy, but by directly reducing DNA lesional burden and improving cell survival. Ex-vivo analysis of aged immune cells from healthy donors revealed a stark enrichment of markers for DNA damage, senescence, and mTORC hyperactivation, suggesting that human immune ageing may be amenable to intervention by low-dose mTOR inhibition. To test this in vivo, we conducted a placebo-controlled experimental medicine trial in older adults administered with low-dose rapamycin. p21, a marker of DNA damage-induced senescence, was significantly reduced in immune cells from the rapamycin compared to placebo group. These findings reveal a previously unrecognised role for mTOR inhibition: direct genoprotection. This mechanism may help explain rapamycins exceptional geroprotective profile and opens new avenues for its use in contexts where genome instability drives pathology, ranging from healthy ageing, clinical radiation exposure, and even the hazards of cosmic radiation in space travel.

immunology↗

bioRxiv2025

EssSubgraph improves performance and generalizability of mammalian essential gene prediction with large networks

Predicting essential genes is important for understanding the minimal genetic requirements of organisms, identifying disease-associated genes, and discovering potential drug targets. Wet-lab experiments for identifying essential genes are time-consuming and labor-intensive. Although various machine learning methods have been developed for essential gene prediction, both systematic testing with large collections of gene knockout data and rigorous benchmarking for efficient methods are very limited to date. Furthermore, current graph-based approaches require learning the entire gene interaction networks, leading to high computational costs, especially for large-scale networks. To address these issues, we propose EssSubgraph, an inductive representation learning method that integrates graph-structured network data with omics features for training graph neural networks. We used comprehensive lists of human essential genes distilled from the latest collection of knockout datasets for benchmarking. When applied to essential gene prediction with multiple types of biological networks, EssSubgraph achieved superior performance compared to existing graph-based and other models. The performance is more stable than other methods with respect to network structure and gene feature perturbations. Because of its inductive nature, EssSubgraph also enables predicting gene functions using dynamical networks with unseen nodes and it is scalable with respect to network sizes. Finally, EssSubgraph has better performance in cross-species essential gene prediction compared to other methods. Our results show that EssSubgraph effectively combines networks and omics data for accurate essential gene identification while maintaining computational efficiency. The source code and datasets used in this study are freely available at https://github.com/wenmm/EssSubgraph.

bioinformatics↗

bioRxiv2025

Transcriptional response to Phytophthora root rot in raspberry identifies RiABP19, a Germin-like protein (GLP) gene with a putative role in resistance

Most phytophthora root rot (PRR) outbreaks in symptomatic commercially cultivated raspberry varieties are associated with the prevalence of Phytophthora rubi. Reduced availability of chemical actives and the persistent presence of Phytophthora oospores in the soil contribute to its devastating impact on raspberry-growing regions. In this study, we examined the variation in root morphology in two contrasting raspberry cultivars, Latham (PRR resistant) and Glen Moy (PRR susceptible). We performed RNA-sequencing on Latham roots challenged with P. rubi, to study the transcriptomic response and uncover mechanisms underpinning resistance. We established a new raspberry reference transcript dataset that allowed quantification of raspberry root gene expression. Transcripts significantly upregulated in Latham challenged with P. rubi, included many with characterised roles in resistance, such as Pathogenesis-related proteins and a Germin-like protein, designated RiABP19. The homologous Glen Moy RiABP19 gene showed no differential transcriptional response to PRR infection, indicating a resistance cultivar-specific induction signature following PRR challenge. Three-dimensional structural modelling predicts that RiABP19 contains conserved active sites implicated in auxin-binding and superoxide dismutase activity and can form a homo-hexamer like true germins. Co-immunoprecipitation assays confirmed that RiABP19 can form both homo- and heterodimers in planta. Virus-induced gene silencing of RiABP19 orthologs of in the model plant Nicotiana benthamiana strongly impacts immune signalling, enhancing Phytophthora infestans colonization and attenuating resistance and cell death triggered by the tomato Cf4/Avr4 interaction. These findings suggest that RiABP19 functions as a positive regulator of immunity and may represent a target for future crop improvement in raspberries.

plant biology↗

bioRxiv2025

Integrating Artificial Intelligence-Driven Digital Pathology and Genomics to Establish Patient-Derived Organoids as a Novel Alternative Model for Drug Response in Head and Neck Cancer

Patient-derived organoids (PDOs) are emerging as advanced 3D ex vivo novel alternative method (NAM) preclinical models, offering significant advantages over traditional cell lines and monolayer cultures for therapeutic development. In this study, we established PDOs from surgically resected fresh tissues of human papillomavirus (HPV)-negative head and neck squamous cell carcinoma (HNSCC) across anatomical sites, tumor T-categories, and sample types. These PDOs faithfully recapitulate the tumors pathology, mutational profile, and drug response. To enable rapid classification of PDO identity, we developed a new convolutional neural network (CNN) model, TransferNet-PDO, which accurately distinguished tumor versus normal PDOs in culture using digital histopathology images (AUC[≥]0.88). PDOs maintained stable cultures and were cryopreserved between passages 5 and 12. Immunohistochemistry (IHC) staining (PanCK, p63, Cytokeratin 13, Ki67) confirmed squamous phenotype and histologic aggression of the original tumor. For tumors harboring TP53 mutations by whole-exome sequencing (WES), PDOs retained the corresponding p53 functional status as confirmed by IHC (enhanced or loss of expression). Somatic mutational landscape revealed that PDOs preserved driver somatic mutations, copy number variations (CNVs), and clonal architecture including low-prevalence subclones. Drug sensitivity assessment of PDOs showed that cisplatin reduced cell viability, whereas cetuximab and lenvatinib had minimal effects. Chemoradiation led to greater tumor organoid killing compared to radiation or chemotherapy alone. This study presents an integrated HNSCC PDO platform combining tissue biobanking, organoid establishment, multi-omics characterization, functional drug screening, and AI-driven histopathologic classification, providing a comprehensive and scalable system for translational cancer research.

cancer biology↗

bioRxiv2025

Proteomic and metabolic profiling reveals APOE4-dependent shifts in whole brain, neuronal, and astrocytic mitochondrial function and glycolysis

Apolipoprotein E (APOE) genetic variation is the strongest genetic risk factor for late onset Alzheimers disease (LOAD). Studies on APOE genotype dependent changes have largely focused on amyloid beta (A{beta}) aggregation, disease pathology, and lipid metabolism. Recently, there has been increased interest in the relationship between metabolic function and APOE genetic variation. In this study, we examined how APOE genotype can alter metabolism in the brains of young male and female APOE3 and APOE4 targeted replacement (TR) mice. In combination with this, we also examined cell type-specific differences using induced pluripotent stem cell (iPSC) derived astrocytes and neurons. We found sex and genotype dependent changes to metabolism in the brains of young APOE TR mice. Specifically, APOE4 mice show signs of metabolic stress and compensatory mechanisms in the brain. Using proteomics and stable isotope tracing metabolomics, we found that APOE4 iAstrocytes and iNeurons exhibit signs of inflammation, mitochondrial dysfunction, altered TCA cycle and malate-aspartate shuttle activity, and a metabolic shift toward glycolysis. Taken together, this data indicates APOE4 causes early changes to metabolism within the central nervous system. While this study establishes a relationship between APOE genotype and alterations in bioenergetics, additional studies are needed to investigate underlying mechanisms.

neuroscience↗

bioRxiv2025

The 18S rRNA methyltransferase, BUD23, is required for appropriate lipid and mitochondrial metabolism

Efficient energy metabolism is essential for health and its dysregulation drives cardiometabolic disease. Delivery of regulatory control through translation and ribosome function is emerging as important. Here, we identify the rRNA methyltransferase BUD23 as a potent regulator of cellular and systemic energy homeostasis. Adipocyte-specific deletion of BUD23 in mice regulates lipid and mitochondrial metabolism resulting in a pronounced lean phenotype and resistance to diet-induced obesity. Mechanistically, BUD23 modulates translation initiation and efficiency of mRNAs with specific features - including short 5 UTR length and GC-rich post-initiation codon usage - characteristic of mitochondrial and lipogenic proteins. Genetic analyses and Mendelian randomisation support a role for BUD23 in human cardiometabolic traits and disease burden. Together, our findings uncover a conserved translational control mechanism that regulates energy state, from cellular metabolism through to human cardiometabolic health.

cell biology↗

bioRxiv2025

Phylogeny-driven design of broadly protective sarbecovirus receptor-binding domain nanoparticle vaccines

Vaccines against emerging SARS-CoV-2 variants and sarbecoviruses with pandemic potential must elicit a robust humoral immune response in a population imprinted with the SARS-CoV-2 spike (S) protein. Here, we designed protein nanoparticle (NP) vaccines co-displaying the SARS-CoV-2 BA.5, SARS-CoV-1, and BtKY72 receptor-binding domains (RBDs) with or without the Wuhan-Hu-1 (Wu) RBD. We show that these vaccines elicit cross-reactive and broadly neutralizing plasma antibody responses against SARS-CoV-2 variants and sarbecoviruses in naive and pre-immune animals. Immunization with multivalent RBD-NPs overcomes immune imprinting and elicits neutralizing antibodies and memory B cells specific for the BA.5, SARS-CoV-1, and BtKY72 RBDs in mRNA-1273-vaccinated non-human primates. Multivalent RBD-NPs outperform a monovalent Wu RBD-NP vaccine by providing superior protection in mice and non-human primates challenged with the vaccine-mismatched SARS-CoV-2 XBB.1.5 or the pre-emergent RsSHC014. These data support the use of multivalent RBD-NP vaccines for SARS-CoV-2 variants and sarbecoviruses in naive and pre-immune populations.

immunology↗

bioRxiv2025

Sex, Stress and the Heart: Long-term Cardiovascular Effects of Embryonic Metabolic Disruption

Adverse conditions within the embryonic environment can alter embryogenesis, programming systemic physiological changes that may manifest as disease states in adult life. The process of developmental programming represents an important factor underlying cardiometabolic diseases, many of which are leading causes of death globally. Importantly, there is evidence that males are less tolerant to certain environmental perturbations during embryogenesis, mirrored by sex differences in the incidence of certain cardiometabolic diseases. Understanding sex differences in programmed responses in mammalian models is complicated by maternal compensation and placental factors. Avian models offer a valuable comparable system in which such effects are not present. Here, we investigate the influence of developmental hypoxia and hypothermia in programming cardiovascular structure and function in the domestic chicken (Gallus gallus domesticus). In agreement to mammalian studies, adult males but not females show pathological mitochondrial morphology and respiratory capacity, ventricular hypertrophy and reduced body weight programmed by embryonic hypothermia and hypoxia. These data not only represent novel findings in birds but demonstrate the utility of the avian model for understanding sex differences in prenatal stress responses, revealing common responses among endothermic amniotes.

physiology↗

bioRxiv2025

Characterization of Clinical Fusobacterium nucleatum Isolates from Oral Squamous Cell Carcinoma Patients

Fusobacterium nucleatum (Fn), a Gram-negative anaerobe primarily residing in the oral cavity, has garnered increasing attention for its role in a broad spectrum of human diseases. While typically absent or rarely detected outside the oral cavity in healthy individuals, Fn is frequently found at extra-oral sites under disease conditions and has been implicated in cancer progression and prognosis1. In oral squamous cell carcinoma (OSCC), the abundance of Fn significantly increases as the disease progresses, promoting cell invasion and metastasis. Furthermore, substantial evidence links Fn to accelerated tumor growth and metastatic progression in colorectal cancer (CRC), where its presence is also associated with chemotherapy resistance and poor prognosis. Here, to further elucidate the pathogenic mechanisms of Fn in cancer progression, this study characterized the physiological traits, virulence factor expression, and impacts on cancer cells of 10 Fn strains, including two well-characterized ATCC strains and eight clinical isolates. The clinical isolates consisted of three strains from saliva samples of 117 OSCC patients and five strains from 160 non-cancer individuals. Results indicated that oral isolates, regardless of disease origin, all belong to Fn subspecies polymorphum. The ATCC strains (23726 and 25586) exhibited shorter cell lengths and faster growth rates compared to the clinical isolates. Both ATCC strains formed stable biofilms and expressed key virulence genes, including aim1, fadA, fomA, and radD. All isolates tested showed sensitivity to a panel of eight different antibiotics. Interestingly, only one clinical isolate displayed similar stimulation of CRC cell migration as the two ATCC strains, while the other seven displayed no such ability. Collectively, these findings suggest that the most virulent strains, in terms of biofilm formation and virulence gene expression, are not necessarily the most pathogenic in the context of cancer cell interactions. Future pan-genomic analyses, incorporating whole-genome sequencing of the clinical isolates, will aim to delineate the genetic determinants contributing to the carcinogenic potential of Fn.

microbiology↗

bioRxiv2025

High-resolution digital dissociation of brain tumors with deep multimodal autoencoder

In single cell biology, the complexity of tissues may hinder lineage cell mapping or tumor microenvi-ronment decomposition, requiring digital dissociation of bulk tissues. Many deconvolution methods focus on transcriptomic assay, not easily applicable to other omics due to ambiguous cell markers and reference-to-target difference. Here, we present MODE, a multimodal autoencoder pipeline linking multi-dimensional features to jointly predict personalized multi-omic profiles and cellular compositions, using pseudo-bulk data constructed by internal non-transcriptomic reference and external scRNA-seq data. MODE was evaluated through rigorous simulation experiments and real multiomic data from multiple tissue types, outperforming nine deconvolution pipelines with superior generalizability and fidelity.

bioinformatics↗