Biology subjects

Chen, S.

Publications and source records attributed to Chen, S..

At least 343 records · Page 19Linked to original sources

bioRxiv2024

Manipulation of the brown glume and internode 1 gene leads to alterations in lignified tissue coloration, lignification, and pathogen resistance in wheat

Lignin is a crucial component of the cell wall, providing mechanical support and protection against biotic and abiotic stresses. However, little is known about wheat lignin-related mutants and their roles in pathogen defense. In this study, we identified an ethyl methanesulfonate (EMS)-derived Aegilops tauschii mutant named brown glume and internode 1 (bgi1), which exhibits reddish-brown pigmentation in various tissues, including internodes, spikes, and glumes. Using map-based cloning and single nucleotide polymorphism (SNP) analysis, we identified AET6Gv20438400 (BGI1) as the leading candidate gene, encoding the TaCAD1 protein. The mutation occurred in the splice acceptor site of the first intron, resulting in a premature stop codon in BGI1. We validated the function of BGI1 using loss-of-function EMS and gene editing knockout mutants, both of which displayed reddish-brown pigmentation in lignified tissues. BGI1 knockout mutants exhibited reduced lignin content and shearing force relative to wild type, while BGI1 overexpression transgenic plants showed increased lignin content and enhanced disease resistance against common root rot and Fusarium crown rot. We confirmed that BGI1 exhibits CAD activity both in vitro and in vivo, playing an important role in lignin biosynthesis. BGI1 was highly expressed in the stem and spike and localized in the cytoplasm. Transcriptome analysis revealed the regulatory networks associated with BGI1. Finally, we demonstrated that BGI1 interacts with TaPYL-1D, potentially involved in the abscisic acid signaling pathway. The identification and functional characterization of BGI1 significantly advance our understanding of CAD proteins in lignin biosynthesis and plant defense against pathogen infection in wheat.

plant biology↗

bioRxiv2024

Ligand-Receptor Interactions for Cell-Cell Communication Analysis in Rat, Chicken, Pig, and Monkey Single-Cell and Spatial Transcriptomics

Cell-cell communication is a frequently used analysis approach in single cell RNA and spatial transcriptomics, and many tools like CellPhoneDB, CellChat and stLearn have been developed. Ligand-receptor interactions are the core of cell-cell communication analysis. Since receptor-ligand and even protein-protein interactions were focus on humans and mice research, curated human and mouse receptor-ligand databases have been established, cell-cell interactions for these two species single-cell RNA sequencing data can be directly analyzed. However, for rats, chickens, pigs, monkeys, and other species, cell-cell interaction analysis is often implemented through orthologous gene mapping, due to the lack of curated ligand-receptor interaction databases for these species. We collected cell-cell interaction data mainly from KEGG for rats, chickens, pigs, and monkeys, and extended the data from Reactome and IntAct. Then, by using CellChatV2 with our collected rat ligand-receptor interactions and CellChatV2s own mouse data, respectively, we analyzed 10x Genomics rat public scRNA data, and found that 70 significantly ligand-receptor interactions from the mouse analysis result were also significantly in rats. We also obtained some chicken, pig, and monkey scRNA data from published literature, and analyzed cell-cell interactions using our collected ligand-receptor interactions for these species, and it was proved that our data is reliable and useful. Lastly, we have transformed the ligand-receptor interactions for rat, chicken, pig, and monkey species into the CellChatDB format, which enables swift and straightforward analysis of cell-cell communication in single-cell and spatial data of these four species. All the ligand-receptor interaction datasets for rats, chickens, pigs, and monkeys, as well as the program codes, are available at https://github.com/qingchen36/ligand-receptor. Using our program, one can rapidly obtain receptor-ligand interaction data for other species.

bioinformatics↗

bioRxiv2024

Nanoparticles use magnetoelectricity to target and eradicate cancer cells

This study presents the first in vivo and in vitro evidence of an externally controlled, predictive, MRI-based nanotheranostic agent capable of cancer cell specific targeting and killing via irreversible electroporation (IRE) in solid tumors. The rectangular-prism-shaped magnetoelectric nanoparticle is a smart nanoparticle that produces a local electric field in response to an externally applied magnetic field. When externally activated, MENPs are preferentially attracted to the highly conductive cancer cell membranes, which occurs in cancer cells because of dysregulated ion flux across their membranes. In a pancreatic adenocarcinoma murine model, MENPs activated by external magnetic fields during magnetic resonance imaging (MRI) resulted in a mean three-fold tumor volume reduction (62.3% vs 188.7%; P < .001) from a single treatment. In a longitudinal confirmatory study, 35% of mice treated with activated MENPs achieved a durable complete response for 14 weeks after one treatment. The degree of tumor volume reduction correlated with a decrease in MRI T2* relaxation time (r = .351; P = .039) which suggests that MENPs have a potential to serve as a predictive nanotheranostic agent at time of treatment. There were no discernable toxicities associated with MENPs at any timepoint or on histopathological analysis of major organs. MENPs are a noninvasive alternative modality for the treatment of cancer. SummaryWe investigated the theranostic capabilities of magnetoelectric nanoparticles (MENPs) combined with MRI via a murine model of pancreatic adenocarcinoma. MENPs leverage the magnetoelectric effect to convert an applied magnetic field into local electric fields, which can induce irreversible electroporation of tumor cell membranes when activated by MRI. Additionally, MENPs modulate MRI relaxivity, which can be used to predict the degree of tumor ablation. Through a pilot study (n=21) and a confirmatory study (n=27), we demonstrated that, [≥]300 {micro}g of MRI-activated MENPs significantly reduced tumor volumes, averaging a three-fold decrease as compared to controls. Furthermore, there was a direct correlation between the reduction in tumor T2 relaxation times and tumor volume reduction, highlighting the predictive prognostic value of MENPs. Six of 17 mice in the confirmatory studys experimental arms achieved a durable complete response, showcasing the potential for durable treatment outcomes. Importantly, the administration of MENPs was not associated with any evident toxicities. This study presents the first in vivo evidence of an externally controlled, MRI-based, theranostic agent that effectively targets and treats solid tumors via irreversible electroporation while sparing normal tissues, offering a new and promising approach to cancer therapy.

bioengineering↗

bioRxiv2024

Mechanical properties are tuned during development with the fibrous network nature of the Arabidopsis cell wall

During plant development, the mechanical properties of the cell walls must be tuned to regulate the growth of the cells. Cell growth involves significant stretching of the cell walls, yet little is known about the mechanical properties of cell walls under such substantial deformation, or how these mechanical properties change to regulate development. Here, we investigated the mechanical behavior of the Arabidopsis leaf epidermal cells being stretched. We found that the mechanical properties arise from the cell wall, which behaves as a fibrous network material. The epidermis exhibited a non-linear stiffening behavior that fell into three regimes. Each regime corresponded to distinct nonlinear behaviors in terms of transverse deformation (i.e., Poisson effect) and unrecoverable deformation (i.e., plasticity). Using a model, we demonstrated that the transition from reorientation and bending-dominated to stretch-dominated deformation modes of cellulose microfibrils cause these nonlinear behaviors. We found the stiffening behavior is more pronounced at later developmental stages. Finally, we show the spiral2-2 mutant has anisotropic mechanical properties, likely contributing to the spiraling of leaves. Our findings reveal the fibrous network nature of cell walls gives a high degree of tunability in mechanical properties, which allows cells to adjust these properties to support proper development.

plant biology↗

bioRxiv2024

Inferring Gene Regulatory Network Based on scATAC-seq Data with Gene Perturbation

Gene regulatory networks (GRNs) are critical blueprints for understanding gene regulation and the intricate interactions that drive biological processes. Recent advances have highlighted the potential of single-cell ATAC-seq (scATAC-seq) data in GRN inference, offering unprecedented insights into how chromatin accessibility plays an important part in gene regulation. However, existing methods often fall short in providing a quantitative and holistic depiction of regulatory relationships, particularly in capturing the strength, direction, and type of gene regulation simultaneously. In this paper, we present a novel approach that addresses these limitations by leveraging genetically perturbed scATAC-seq data to infer more comprehensive and accurate GRNs. Our method advances the field by integrating pre- and post-perturbation chromatin accessibility data, enabling the construction of GRNs that more accurately reflect the dynamic regulatory landscape. Through rigorous evaluation on seven real datasets, we demonstrate the methods superior performance in reconstructing GRNs with enhanced precision and interpretability. This work significantly contributes to the field by providing a robust framework for GRN inference, with broad implications for understanding gene regulation in complex biological systems.

bioinformatics↗

bioRxiv2024

A Suite of Foundation Models Captures the Contextual Interplay Between Codons

In the canonical genetic code, many amino acids are assigned more than one codon. Work by us and others has shown that the choice of these synonymous codon is not random, and carries regulatory and functional consequences. Existing protein foundation models ignore this context-dependent role of coding sequence in shaping the protein landscape of the cell. To address this gap, we introduce cdsFM, a suite of codon-resolution large language models, including both EnCodon and DeCodon models, with up to 1B parameters. Pre-trained on 60 million protein-coding sequences from more than 5,000 species, our models effectively learn the relationship between codons and amino acids, recapitualing the overall structure of the genetic code. In addition to outperforming state-of-the-art genomic foundation models in a variety of zero-shot and few-shot learning tasks, the larger pre-trained models were superior in predicting the choice of synonymous codons. To systematically assess the impact of synonymous codon choices on protein expression and our models ability to capture these effects, we generated a large dataset measuring overall and surface expression levels of three proteins as a function of changes in their synonymous codons. We showed that our EnCodon models could be readily fine-tuned to predict the contextual consequences of synonymous codon choices. Armed with this knowledge, we applied EnCodon to existing clinical datasets of synonymous variants, and we identified a large number of synonymous codons that are likely pathogenic, several of which we experimentally confirmed in a cellbased model. Together, our findings establish the cdsFM suite as a powerful tool for decoding the complex functional grammar underlying the choice of synonymous codons.

genomics↗

bioRxiv2024

CREATE: cell-type-specific cis-regulatory elements identification via discrete embedding

Identifying cis-regulatory elements (CREs) within non-coding genomic regions--such as enhancers, silencers, promoters, and insulators--is pivotal for elucidating the intricate gene regulatory mechanisms underlying complex biological traits. The current prevalent sequence-based methods often focus on singular CRE types, limiting insights into cell-type-specific biological implications. Here, we introduce CREATE, a multimodal deep learning model based on the Vector Quantized Variational AutoEncoder framework, designed to extract discrete CRE embeddings and classify multiple CRE classes using genomic sequences, chromatin accessibility, and chromatin interaction data. CREATE excels in accurate CRE identification and exhibits strong effectiveness and robustness. We showcase CREATEs capability in generating comprehensive CRE-specific feature spectrum, offering quantitative and interpretable insights into CRE specificity. By enabling large-scale prediction of CREs in specific cell types, CREATE facilitates the recognition of disease- or phenotype-related biological variabilities of CREs, thereby expanding our understanding of gene regulation landscapes.

bioinformatics↗

bioRxiv2024

Effect heterogeneity reveals complex pleiotropic effects of rare coding variants

Recent expansion of large-scale biobank resources has enabled rare-variant association studies (RVAS) and systematic investigation of rare variant pleiotropy across thousands of phenotypes simultaneously. However, existing statistical frameworks for dissecting pleiotropy were largely developed for common variants and are not well suited to gene-based rare variant signals, limiting the interpretation of cross-phenotype associations. Here, we develop ALLSPICE, a likelihood-based method that tests whether rare variant effects in genes exhibiting cross-phenotype associations are proportional or heterogeneous across continuous traits while accounting for phenotypic correlation using summary statistics. ALLSPICE is well-calibrated in simulations and is implemented as an R package for scalable analysis of gene-level rare-variant burden associations. We applied ALLSPICE to RVAS of 359 continuous traits in the UK Biobank and identified 124 significant heterogeneous events among 11,810 pairs of gene-trait associations. By identifying effect heterogeneity within genes associated with multiple phenotypes, ALLSPICE clarifies shared and heterogeneous rare variant architectures underlying cross-phenotype associations and provides insight into rare variant pleiotropy.

genetics↗

bioRxiv2024

An evolutionarily conserved Hox-Gbx segmentation code in the rice coral Montipora capitata

Segmentation of the gastric cavity is a synapomorphic trait of cnidarians of the class Anthozoa (corals and sea anemones), with different clades forming distinct numbers of segments. In the starlet sea anemone Nematostella vectensis, for example, eight bilaterally positioned gastric segments are generated by the action of a group of Hox-Gbx genes in the developing larval endo-mesoderm. Still, given the range of segment numbers observed in different anthozoans, it remains unclear whether this Hox-Gbx module is evolutionarily conserved and how it might be deployed to generate different numbers of segments. Here, we systematically interrogate the role of Hox-Gbx genes during development of the rice coral Montipora capitata. We first characterize the temporal sequence of segmentation in M. capitata juveniles and then combine transcriptomic profiling and in situ hybridization to identify three conserved homeobox-containing genes, McAnthox8, McAnthox6a.1 and McGbx, which are collectively expressed in the developing endo-mesoderm prior to and during segment formation. The expression boundaries of these genes prefigure the positions of the first six segment boundaries, similar to their Nematostella homologs. Further, we show that chemical inhibition of BMP activity at the planula stage abolishes the expression of Hox-Gbx genes, leading to the formation of an unsegmented gastric cavity. These findings demonstrate the existence of a functionally conserved Hox-Gbx module in evolutionarily divergent anthozoan species, suggesting that the last common ancestor of all anthozoans likely utilized a similar genetic toolkit to axially pattern the endo-mesoderm into metameric subunits.

evolutionary biology↗

bioRxiv2024

Building a literature knowledge base towards transparent biomedical AI

As artificial intelligence (AI) continues to advance and scale up in biomedical research, concerns about AIs trustworthiness and transparency have grown. There is a critical need to systematically bring accurate and relevant biomedical knowledge into AI applications for transparency and provenance. Knowledge graphs have emerged as a powerful tool that integrates heterogeneous knowledge by explicitly describing biomedical knowledge as entities and relationships between entities. However, PubMed, the largest and most comprehensive repository of biomedical knowledge, exists primarily as unstructured text and is under utilized for advanced machine learning tasks. To address the challenge, we developed LiteralGraph, a computational framework to extract biomedical terms and relationships from PubMed literature into a unified knowledge graph. Using this framework, we established the Genomic Literature Knowledge Base (GLKB), which consolidates 14,634,427 biomedical relationships between 3,276,336 biomedical terms from over 33 million PubMed abstracts and nine well-established biomedical repositories. The database is coupled with RESTful APIs and a user-friendly web interface that makes it accessible to researchers for various usages. We demonstrated the broad utility of GLKB towards transparent AI in three distinct application scenarios. In the LLM grounding scenario, we developed a Retrieval Augmented Generation (RAG) agent to reduce LLM hallucination in biomedical question answering. In the hypothesis generation scenario, we elucidated the potential functions of RFX6 in type 2 diabetes (T2D) using the vast evidence from PubMed articles. In the machine learning scenario, we utilized GLKB to provide semantic knowledge in predictive tasks and scientific fact-checking.

bioinformatics↗

bioRxiv2024

Distinct duration and diet dependent lipid profiles and renal complicationsin GCK-MODY

Heterozygous inactivating mutations in the glucokinase (GCK) gene cause one of the most common types of maturity-onset diabetes of the young type 2 (MODY2), also named GCK-MODY. Studies suggest that, unlike other types of diabetes, patients with GCK-MODY do not have increased risk of diabetic complications and therefore do not typically require antihyperglycemic therapy. However, long-term outcomes of GCK-MODY on lipid profiles and chronic complications under different dietary scenarios remain unclear. Herein, using a knock-in mouse model expressing a novel MODY causing mutation GCK-Q26L (GCKMut), we examined age- and diet-related lipid profiles and diabetic complications. We found that, although GCKMut mice exhibited mild elevated blood glucose, the lipid profiles, body fat composition, and diabetic kidney disease (DKD) were initially improved in high-fat-diet (HFD) fed mice at the age of 28 weeks, supporting potential beneficial effects of GCKMut on lipid metabolism and kidney health. Unexpectedly however, those protective effects diminished by 40 weeks, and became more severe dyslipidemia and kidney injury associated with renal fibrosis and inflammation at 60-week-old mice fed with normal diet (ND) or HFD. Those findings revealed distinct duration- and diet-dependent effects of inactivating GCK mutation on lipid profiles and DKD, highlighting previously unrealized long-term chronic complications in GCK-MODY.

physiology↗

bioRxiv2024

The Precise Basecalling of Short-Read Nanopore Sequencing

The nanopore sequencing of short sequences, whose lengths are typically less than 0.3kb therefore comparable with Illumina sequencing techniques, has recently gained wide attention. Here, we design a scheme for training nanopore basecallers that are specialized for short biomolecules. With bioengineered RNA (BioRNA) molecules as examples, we demonstrate the superior accuracy of basecallers trained by our scheme.

bioinformatics↗

bioRxiv2024

Islet amyloid disrupts MHC Class II antigen presentation and protects NOD mice from autoimmune diabetes.

Islet amyloid contributes to beta cell failure in type 2 diabetes through several mechanisms, one being the potent induction of local islet inflammation through activating inflammatory pathways in islet macrophages. We performed an unbiased phenotypic investigation of islet macrophages in the early stage of islet amyloid formation using single cell RNA sequencing of resident islet macrophages in mice with and without the amyloidogenic form of human islet amyloid polypeptide (hIAPP). This revealed that MHC Class II antigen presentation genes were strongly down-regulated in islet macrophages during islet amyloid formation. As islet amyloid has recently been reported in pancreases of people with type 1 diabetes, we sought to investigate the impact of islet amyloid in the NOD mouse model of type 1 diabetes. Both overexpression and physiological expression of hIAPP delayed diabetes in NOD mice relative to littermate controls, corresponding with decreased markers of antigen presentation and activation, as well as decreased immune cell infiltration in islets. Adoptive transfer studies showed that systemic autoimmune function remained intact and beta cells from hIAPP transgenic mice did not evade immune recognition by diabetogenic T cells, collectively indicating the protection from diabetes was mediated by localized disruption of antigen presentation in the pancreas. Consistent with this, incubation of dendritic cells with IAPP aggregates decreased MHC Class II surface expression and diminished antigen-specific T cell activation in vitro, through a phagocytosis-dependent mechanism. Collectively our data show that despite the well-established pro-inflammatory response of macrophages to IAPP aggregates, the uptake of IAPP aggregates during early amyloid formation also disrupts MHC Class II antigen presentation and slows beta cell autoimmunity.

immunology↗

bioRxiv2024

The Arabidopsis histone H3K4me3-binding ALFIN-like proteins mediate histone H2A ubiquitination and coordinate diverse chromatin modifications

The histone H3K4 trimethylation (H3K4me3) is widely distributed at numerous actively transcribed protein-coding genes throughout the genome. However, the interplay between H3K4me3 and other chromatin modifications remains poorly understood in plants. In this study, we find that the Arabidopsis thaliana H3K4me3-binding ALFIN-LIKE (AL) proteins are associated with H3K4me3-enriched genes at the whole-genome level. The AL proteins contain a C-terminal PHD finger, which has a conserved role in recognizing H3K4me3, and a PHD-associated AL (PAL) domain, which is responsible for binding to diverse chromatin-related proteins. We demonstrate that the AL proteins not only act as subunits of the Polycomb repressive complex 1 (PRC1) to mediate H2A ubiquitination at H3K4me3-enriched genes but also interact with a variety of other chromatin-related proteins. Furthermore, we elucidate the mechanisms by which AL proteins interact with other chromatin-associated proteins to integrate H3K4me3, H2A ubiquitination, H2A.Z deposition, H3K27 demethylation, and chromatin accessibility across the genome. These findings underscore the critical role of AL proteins in linking H3K4me3 with a variety of other chromatin modifications in plants.

plant biology↗

bioRxiv2024

EPOP Restricts PRC2.1 Targeting to Chromatin by Directly Modulating Enzyme Complex Dimerization

Polycomb repressive complex 2 (PRC2) mediates developmental gene repression as two classes of holocomplexes, PRC2.1 and PRC2.2. EPOP is an accessory subunit specific to PRC2.1, which also contains PCL proteins. Unlike other accessory subunits that collectively facilitate PRC2 targeting, EPOP was implicated in an enigmatic inhibitory role, together with its interactor Elongin BC. We report an unusual molecular mechanism whereby EPOP regulates PRC2.1 by directly modulating its oligomerization state. EPOP disrupts the PRC2.1 dimer and weakens its chromatin association, likely by disabling the avidity effect conferred by the dimeric complex. Congruently, an EPOP mutant specifically defective in PRC2 binding enhances genome-wide enrichments of MTF2 and H3K27me3 in mouse epiblast-like cells. Elongin BC is largely dispensable for the EPOP-mediated inhibition of PRC2.1. EPOP defines a distinct subclass of PRC2.1, which uniquely maintains an epigenetic program by preventing the over-repression of key gene regulators along the continuum of early differentiation.

molecular biology↗

bioRxiv2024

Impact of Acetate and Optimized Nitrate Levels on Mixotrophic Growth and Protein Dynamics in Chlorella Sorokiniana

Microalgae are well-known for their role as sustainable bio-factories, offering a promising solution to the global food and nutrition crisis. To clarify the potential of Chlorella sorokiniana UTEX 1230 for food applications, particularly as an alternative protein source, the study employed a mixotrophic cultivation mode with sodium acetate (NaAc) as a cost-effective organic carbon (NaAc-C) source. Varying levels of NaAc-C and nitrate-sourced nitrogen were investigated, optimizing the effect of metabolic characteristics of the microalgal growth. The designed heterotrophic cultivation confirmed the ability of C. sorokiniana UTEX 1230 to grow on NaAc-C, and then the mixotrophic cultures, when supported by both NaAc-C and CO2, exhibited superior growth performance, achieving double the biomass concentration compared to the autotrophic control. The addition of nitrogen (750 mg/L NaNO) facilitated the thorough metabolism of NaAc-C and enhanced photosynthetic activity indicated by a 196% increase in pigment levels, which resulted in a maximum biomass concentration of 2.82 g/L in the 150 mM NaAc-C group. A detailed analysis of nitrogen and protein concentrations over time revealed that higher nitrogen availability led to greater protein accumulation which was then degraded to support essential life activities under nitrogen starvation. Therefore, it is suggested that supplementing nitrate on the 3rd day and harvesting on the 4th day could be strategically implemented to increase protein yield from 0.17 g/L/d to 0.34 g/L/d. These findings offer theoretical guidance for further refining this microalgal strain for use as an alternative protein.

bioengineering↗

bioRxiv2024

Endogenous retroviruses activate MARCO-mediated inflammatory response to block retroviral infection

Endogenous retroviruses (ERVs) are remnants of ancient retroviral infections and can profoundly affect the host antiviral innate immune response, although the mechanisms by which these changes occur are largely unknown. Here we report that chicken-specific ERVs exert genetic resistance to exogenous retrovirus infection. Mechanistically, chicken-specific ERVs activated the scavenger receptor MARCO (macrophage receptor with collagenous structure)-mediated TLR3-IL-1{beta} inflammatory response in macrophages. Under the presence of MARCO, macrophages response to viral infection through inducing TLR3-IL-1{beta} inflammatory response. Conversely, lack of MARCO increased the viral replication levels and attenuated the antiviral inflammatory response. MARCO-mediated ligand delivery enhances TLR3-IL-1{beta} antiviral response, and IL-1{beta} expression is responsible for viral inhibition. Restoring MARCO or IL-1{beta} expression overcomes viral infection in macrophages. Our study provides new insights into the molecular mechanisms underlying the host defense against exogenous retroviruses infection and may have important implications for the development of novel therapeutic strategies against retroviruses infection.

immunology↗

bioRxiv2024

RNase III in Salmonella Enteritidis enhances bacterial virulence by reducing host immune responses

Salmonella is an important foodborne pathogen which comprises strains that exhibit varied virulence phenotypes and the capability of causing invasive human infection. In this study, the gene expression profile of foodborne and clinical Salmonella strains that exhibit high- and low-level virulence was investigated, with results showing that the expression level of a number of genes, including the rnc gene which encodes the RNase III ribonuclease, were exceptionally high in the high virulence strains. Investigation of the role of rnc in mediating expression of virulence phenotypes in Salmonella showed that the product of this gene could enhance expression of the superoxide dismutase SodA, which is an essential determinant of survival fitness of Salmonella under the oxidative stress elicited by the host immunity. On the other hand, we also discovered that the double-stranded RNA (dsRNA) released from Salmonella could trigger immune response of the host, and that the high-level expression of the rnc gene enabled Salmonella to evade the host immunity by reducing the amount of dsRNA accumulated in the bacterial cell. These findings provide insightful understanding of the regulation of Salmonella virulence and facilitate development of novel antimicrobial treatments through suppression of virulence expression and survival fitness of this important pathogen.

microbiology↗