Biology subjects

Yang, Y.

Publications and source records attributed to Yang, Y..

At least 1,243 records · Page 69Linked to original sources

bioRxiv2020

O-GlcNAcylation of SAMHD1 Indicating a Link between Metabolic Reprogramming and Anti-HBV Immunity

Viruses hijack the host cell machinery to promote viral replication; however, the mechanism by which metabolic reprogramming regulates innate antiviral immunity in the host remains elusive. Herein, we found that Hepatitis B virus (HBV) infection upregulates glucose transporter 1expression, promotes hexosamine biosynthesis pathway (HBP) activity, and enhances O-linked {beta}-N-acetylglucosamine (O-GlcNAc) modification of downstream proteins. HBP-mediated O-GlcNAcylation positively regulates host antiviral response against HBV in vitro and in vivo. Mechanistically, O-GlcNAc transferase (OGT)-mediated O-GlcNAcylation of sterile alpha motif and histidine/aspartic acid domain-containing protein 1 (SAMHD1) on Ser93 stabilizes SAMHD1 and enhances its antiviral activity. In addition, O-GlcNAcylation of SAMHD1 promoted its antiviral activity against human immunodeficiency virus-1 in vitro. In conclusion, the results of our study reveal a link between HBP, O-GlcNAc modification, and innate antiviral immunity by targeting SAMHD1. Therefore, the results of this study demonstrate a strategy for the potential treatment of HBV infection by modulating HBP activity.

microbiology↗

bioRxiv2020

Eukaryotic transposable elements as "cargo carriers": the forging of metal resistance in the fungus Paecilomyces variotii

The horizontal transfer of large gene clusters by mobile elements is a key driver of prokaryotic adaptation in response to environmental stresses. Eukaryotic microbes face similar environmental stresses yet a parallel role for mobile elements has not yet been established. A stress faced by all microorganisms is the prevalence of toxic metals in their environment. In fungi, identified mechanisms for protection against metals generally rely on genes that are dispersed within an organisms genome. Here we have discovered a large ([~]85 kb) region that confers resistance to several metals in the genomes of some, but not all, strains of a fungus, Paecilomyces variotii. We name this region HEPHAESTUS (H{phi}) and present evidence that this region is mobile within the P. variotii genome with features highly characteristic of a transposable element. While large gene clusters including those for the synthesis of secondary metabolites have been widely reported in fungi, these are not mobile within fungal genomes. HEPHAESTUS contains the greatest complement of host-beneficial genes carried by a transposable element in eukaryotes. This suggests that eukaryotic transposable elements might play a role analogous to their bacterial counterparts in the horizontal transfer of large regions of host-beneficial DNA. Genes within HEPHAESTUS responsible for individual metal resistances include those encoding a P-type ATPase transporter, PcaA, required for cadmium and lead resistance, a transporter, ZrcA, providing resistance to zinc, and a multicopper oxidase, McoA, conferring resistance to copper. Additionally, a subregion of H{phi} conferring resistance to arsenate was identified. The presence of a strikingly similar cluster in the genome of another fungus, Penicillium fuscoglaucum, suggests that HEPHAESTUS arrived in P. variotii via horizontal gene transfer.

molecular biology↗

bioRxiv2020

FER-LIKE FE DEFICIENCY-INDUCED TRANSCRIPTION FACTOR (OsFIT) interacts with OsIRO2 to regulate iron homeostasis

There are two Fe-uptake strategies for maintaining Fe homeostasis in plants. As a special graminaceous plant, rice applies both strategies. However, it remains unclear how these two strategies are regulated in rice. IRON-RELATED BHLH TRANSCRIPTION FACTOR 2 (OsIRO2) is critical for regulating Fe uptake in rice. In this study, we identified an interacting partner of OsIRO2, Oryza sativa FER-LIKE FE DEFICIENCY-INDUCED TRANSCRIPTION FACTOR (OsFIT), which encodes a bHLH transcription factor. The OsIRO2 protein is localized in the cytoplasm and nucleus, but OsFIT facilitates the accumulation of OsIRO2 in the nucleus. Loss-of-function mutations to OsFIT result in decreased Fe accumulation, severe Fe-deficiency symptoms, and disrupted expression of Fe-uptake genes. In contrast, OsFIT overexpression promotes Fe accumulation and the expression of Fe-uptake genes. Genetic analyses indicated that OsFIT and OsIRO2 function in the same genetic node. Further analysis suggested that OsFIT and OsIRO2 form a functional transcription activation complex to initiate the expression of Fe-uptake genes. Our findings provide a mechanism understanding of how rice maintains Fe homeostasis. One-sentence summaryOsFIT interacts with and facilitates the accumulation of OsIRO2 in the nucleus where the OsFIT-OsIRO2 transcription complex initiates the transcription of Fe deficiency responsive genes.

plant biology↗

bioRxiv2020

Viral Architecture of SARS-CoV-2 with Post-Fusion Spike Revealed by Cryo-EM

Since December 2019, the outbreak of Coronavirus Disease 2019 (COVID-19) spread from Wuhan, China to the world, it has caused more than 87,000 diagnosed cases and more than 3,000 deaths globally. To fight against COVID-19, we carried out research for the near native SARS-CoV-2 and report here our preliminary results obtained. The pathogen of the COVID-19, the native SARS-CoV-2, was isolated, amplified and purified in a BSL-3 laboratory. The whole viral architecture of SARS-CoV-2 was examined by transmission electron microscopy (both negative staining and cryo-EM). We observed that the virion particles are roughly spherical or moderately pleiomorphic. Spikes have nail-like shape towards outside with a long body embedded in the envelope. The morphology of virion observed in our result indicates that the S protein of SARS-CoV-2 is in post-fusion state, with S1 disassociated. This state revealed by cryo-EM first time could provide an important information for the identification and relevant clinical research of this new coronavirus.

microbiology↗

bioRxiv2020

Inferring Multiple Metagenomic Association Networks based on Variation of Environmental Factors

Identifying significant biological relationships or patterns is central to many metagenomic studies. Methods that estimate association networks have been proposed for this purpose, but they assume that associations are static, neglecting the fact that relationships in a microbial ecosystem may vary with changes in environmental factors, which can result in inaccurate estimations. We propose a computational model, k-Lognormal-Dirichlet-Multinomial model (kLDM), which estimates multiple association networks that correspond to specific environmental conditions according to values of environmental factors (EFs), and simultaneously infers microbe-microbe and environmental factor-microbe associations for each network. We showed the effectiveness of kLDM on synthetic data, a colorectal cancer dataset, as well as the TARA Oceans and American Gut project datasets. The results showed that the widely used Spearmans rank correlation coefficient (SCC) performed much worse than other methods, indicating the importance of separating samples by environmental conditions. We compared cancer fecal samples with cancer-free samples, and our estimation showed fewer associations among microbes but stronger associations between specific bacteria such as five colorectal cancer (CRC)-associated OTUs, indicating gut microbe translocation in cancer patients. Some environmental-factor-dependent associations were found within marine eukaryotic community, and gut microbial heterogeneity of irritable bowel disease (IBD) patients was detected. Results demonstrated that kLDM could successfully unravel the underlying biological associations. In summary, our study presents a computational framework that can elucidate the complex associations within microbial ecosystems. The kLDM program, R, and python scripts, together with all experimental datasets are all accessible at Github (https://github.com/tinglab/kLDM.git).

bioinformatics↗

bioRxiv2020

Children Developing Celiac Disease Have a Distinct and Proinflammatory Gut Microbiota in the First 5 Years of Life

Celiac disease (CD) is an autoimmune disorder triggered by gluten consumption. To identify the role of gut microbes in CD onset, we performed a longitudinal study focusing on two important phases of gut microbiota development at ages 2.5 and 5 (n=16). We obtained samples from children who developed CD during or after the study (CD progressors) and age, sex, and HLA-matched healthy controls. CD progressors had a distinct gut microbiota composition and IgA-sequencing identified unique IgA targets in the gut. Three cytokines, one chemokine, and 19 plasma metabolites were significantly altered in CD progressors at age 5. Feeding C57BL/6J mice with taurodeoxycholic acid (TDCA), a 2-fold increased microbiota-derived metabolite in CD progressors, caused villous atrophy, increased intraepithelial lymphocytes (IELs), CD4+ T-cells, Natural Killer cells, and Qa-1 expression on T-cells while decreasing T-regulatory cells in IELs. Thus, TDCA drives inflammation in the small intestines that potentially contribute to the CD onset. HighlightsO_LICD progressors have a distinct gut microbiome composition compared to healthy controls in two important phases of gut microbiota development (age 2.5 and 5 years) C_LIO_LICD progressors have more IgA-coated bacteria in their gut at age 5 compared to healthy controls. Further, IgA-sequencing identified unique bacterial targets in CD progressors. C_LIO_LIThree plasma proinflammatory cytokines and a chemokine were increased in CD progressors years before diagnosis, indicating an early inflammatory response. C_LIO_LIWe identified 19 metabolites that are significantly altered in CD progress at age 5 and microbiota-derived TDCA increased two-fold. C_LIO_LITDCA treatment in B6 mice increased CD4+ cells and NK cells while decreasing CD8+ T-regulatory (Treg) cells. It also increased Qa-1 expression on immune cells. C_LI

microbiology↗

bioRxiv2020

Alleviating chronic ER stress by p38-Ire1-Xbp1 pathway and insulin-associated autophagy in C. elegans neurons

ER stress occurs in many physiological and pathological conditions. However, how chronic ER stress is alleviated in specific cells in an intact organism is an outstanding question. Here, overexpressing the gap junction protein UNC-9 (Uncoordinated) in C. elegans neurons triggers the Ire1-Xbp1-mediated stress response in an age-dependent and cell-autonomous manner. The p38 MAPK PMK-3 acts upon the phosphorylation of IRE-1 to regulate chronic stress. Overexpressing gap junction protein also activates autophagy and the insulin pathway functions through autophagy, but not the transcription of genes encoding ER chaperones, to counteract the p38-Ire1-Xbp1-mediated stress response. Together, these results reveal an intricate cellular regulatory network in response to chronic stress in a subset of cells in multicellular organism. Author SummaryThe accumulation of unfolded proteins triggers the ER stress response (UPR), which allows cells to fight against fluctuations in protein expression under both physiological and pathological conditions. Severe acute ER stress responses can be induced by drug treatment. However, such intense ER stress rarely occurs ubiquitously in every cell type in vivo. Here, we designed a genetic system in the nematode C. elegans, which allows us to induce ER stress in specific cells, without drug treatment or any other external stimuli, and then to monitor the stress response. Using this system, genetic screens identified that the p38 MAPK directly acts through the IRE-1-XBP-1 branch of the UPR to promote the stress response. Meanwhile, the insulin receptor functions through autophagy activation to counteract the p38-IRE-1-XBP-1 pathway. Together, these results reveal an intricate cellular regulatory network in response to chronic stress in multicellular organism.

genetics↗

bioRxiv2020

Building a Reference Transcriptome for the Hexaploid Hard Fescue Turfgrass (Festuca brevipila) Using a Combination of PacBio IsoSeq and Illumina Sequencing

BackgroundHard fescue (Festuca brevipila Tracey, 2n=6x=42) is a cool season turfgrass with a fine leaf texture that performs well under low-input management. Breeding and genetics studies of F. brevipila have been limited due to the complexity of its hexaploid genome. To advance our knowledge of F. brevipila genomics, we used PacBio isoform sequencing to develop a reference transcriptome for this species. ResultsHere, we report the F. brevipila reference transcriptome generated from root, crown, leaf, and inflorescence tissues. We obtained 59,510 full-length transcripts, of which 38,556 were non-redundant full-length transcripts. The longest and shortest transcripts were 11,487 and 58 bp, respectively. Distribution of synonymous distances among paralogs within F. brevipila suggested highly similar subgenomes that are difficult to distinguish from sequencing errors. To evaluate the phylogenetic relationships among F. brevipila and close relatives, we sequenced three additional transcriptomes using closely related species on an Illumina platform. The results of our phylotranscriptomic analysis supported the close relationships among F. brevipila (6x), Festuca ovina (4x), Festuca ovina subsp. ovina (2x), and Festuca valesiaca (2x), with high levels of discordance among gene trees. ConclusionsOverall, the F. brevipila PacBio Isoseq reference transcriptome provided the foundation for transcriptome studies and allowed breeders a resource for gene discovery in this important turfgrass species.

plant biology↗

bioRxiv2020

Semi-analytical dynamic modeling of DNA surface-hybridization via AC Electro-kinetic steering

The change in electrical property (capacitance) upon hybridization of the desired ssDNA to a capture probe has been proposed as a promising technology platform in biomedical research and practice. An appropriate mathematical model is needed for understanding and optimizing the process occurring at the electrode/electrolyte interface. It is also informative for examining the forces generated by the AC electric fields on the DNA molecules as well as the suspending buffer solution in the experimental pool. Here, we provide the development, formulation and validation of a semi-analytical model of DNA hybridization with deoxynucleotide molecules chemically tethered to a solid gold electrode. The parameters of the proposed model have been estimated using available experimental data. We demonstrate that the detection limit and specificity of our surface-based genosensor are not only dependent on the probe/target binding affinity, but also on the Self-Assembled Monolayer (SAM) density and on the interfacial electric field. The label-free Electrochemical Impedance Spectroscopy (EIS)-based oligonucleotide biosensor with integrated DC-biased can achieve rapid hybridization, high selectivity and sensitive detection for DNA target samples. SIGNIFICANCEDNA hybridization, wherein strands of DNA form duplex through noncovalent, sequence-specific interactions, is one of the most fundamental processes in biology. Fast and reliable determination of miniature amounts of DNA plays important role in clinical forensic and pharmaceutical applications. Thus, developing a better understanding of the kinetic and dynamic properties of DNA hybridization will help in the elucidation of all mechanisms involved in numerous biochemical processes. Moreover, because DNA hybridization has been widely adapted in biotechnology, its study is invaluable to the development of a range of commercially important processes. To achieve optimal sensitivity with minimum sample size and rapid hybridization, ability to predict the kinetics of hybridization based on the characteristics of the strands is crucial, and hence a computer aided numerical model for the design and optimization of a DNA biosensor has been implemented.

biophysics↗

bioRxiv2020

A quantitative framework reveals the ecological drivers of grassland soil microbial community assembly in response to warming

Unraveling the drivers controlling community assembly is a central issue in ecology. Selection, dispersal, diversification and drift are conceptually accepted as major community assembly processes. Defining their relative importance in governing biodiversity is compellingly needed, but very challenging. Here, we present a novel framework to quantitatively infer community assembly mechanisms by phylogenetic bin-based null model analysis (iCAMP). Our results with simulated microbial communities showed that iCAMP had high accuracy (0.93 - 0.99), precision (0.80 - 0.94), sensitivity (0.82 - 0.94), and specificity (0.95 - 0.98), which were 10-160% higher than those from the entire community-based approach. Applying it to grassland microbial communities in response to experimental warming, our analysis showed that homogeneous selection (38%) and "drift" (59%) played dominant roles in controlling grassland soil microbial community assembly. Interestingly, warming enhanced homogeneous selection, but decreased "drift" over time. Warming-enhanced selection was primarily imposed on Bacillales in Firmicutes, which were strengthened by increased drought and reduced plant productivity. This general framework should also be useful for plant and animal ecology.

ecology↗

bioRxiv2020

Lower Soil Carbon Loss Due to Persistent Microbial Adaptation to Climate Warming

Soil microbial respiration is an important source of uncertainty in projecting future climate and carbon (C) cycle feedbacks. Despite intensive studies for two decades, the magnitude, direction, and duration of such feedbacks are uncertain, and their underlying microbial mechanisms are still poorly understood. Here we examined the responses of soil respiration and microbial community structure to long-term experimental warming in a temperate grassland ecosystem. Our results indicated that the temperature sensitivity of soil microbial respiration (i.e., Q10) persistently decreased by 12.0{+/-}3.7% across 7 years of warming. Integrated metagenomic and functional analyses showed that microbial community adaptation played critical roles in regulating respiratory acclimation. Incorporating microbial functional gene abundance data into a microbially-enabled ecosystem model significantly improved the modeling performance of soil microbial respiration by 5-19%, compared to the traditional non-microbial model. Model parametric uncertainty was also reduced by 55-71% when gene abundances were used. In addition, our modeling analyses suggested that decreased temperature sensitivity could lead to considerably less heterotrophic respiration (11.6{+/-}7.5%), and hence less soil C loss. If such microbially mediated dampening effects occur generally across different spatial and temporal scales, the potential positive feedback of soil microbial respiration in response to climate warming may be less than previously predicted.

ecology↗

bioRxiv2020

An original structural fold underlies the multitask P1, a silencing suppressor encoded by the Rice yellow mottle virus

ABSTRACTThe Rice Yellow Mottle sobemovirus (RYMV) belongs to the most damaging pathogens devastating rice fields in Africa. P1, a key protein for RYMV, was reported as a potent RNAi suppressor counteracting RNA silencing in plant reporter systems. Here we describe the complete 3D structure and dynamics of P1. Its N-terminal region contains ZnF1, a structural CCCC-type zinc finger strongly affine to zinc and a prominent short helix, rendering this region poorly amenable to structural changes. P1 C-terminal region contains ZnF2, an atypical HCHC-type ZnF that does not belong to any existing class of Zn finger proteins. ZnF2 appeared much less affine to zinc and more sensitive to oxidizing environments than ZnF1, and may serve as a sensor of plant redox status. The structure helped us to identify key residues essential for RYMV infectivity and spread in rice tissues through their participation in P1 oligomerization and folding. Altogether, our results provide the first complete structure of an antiviral silencing suppressor encoded by a virus infecting rice and highlight P1 structural and dynamical properties that may serve RYMV functions to infect and invade its host plant.Competing Interest StatementThe authors have declared no competing interest.View Full Text

pathology↗

bioRxiv2020

Tracing the origin of a new organ by inferring the genetic basis of rumen evolution

The rumen is the hallmark organ of ruminants and hosts a diverse ecosystem of microorganisms that facilitates efficient digestion of plant fibers. We used 897 transcriptomes from three Cetartiodactyla lineages: ruminants, camels and cetaceans, as well as data from ruminant comparative genomics and functional assays to explore the genetic basis of rumen origin and evolution. Comparative analyses reveal that the rumen and the first-chamber stomachs of camels and cetaceans shared a common tissue origin from the esophagus. The rumen recruited genes from other tissues/organs and up-regulated many esophagus genes to aquire functional innovations involving epithelium absorption, improvement of the ketone body metabolism and regulation of microbial community. These innovations involve such genetic changes as ruminant-specific conserved elements, newly evolved genes and positively selected genes. Our in vitro experiements validate the functions of one enhancer, one positively selected gene and two newly evolved antibacterial genes. Our study provides novel insights into the origin and evolution of a complex organ.

evolutionary biology↗

bioRxiv2020

DGAT-onco: A powerful method to detect oncogenes by integrating differential mutational analysis and functional impacts of somatic mutations

MotivationOncogenes are genes whose malfunctions play critical roles in cancer development, and their discovery is a major aim of cancer mechanisms study. By counting the mutation frequency, oncogenes have been identified with frequent mutations, while it is believed that many more oncogenes could be discovered by differential mutational profile analysis. However, it is common that current methods only utilize mutations in the cancer population, which have an obvious bias in background mutation modelling. MethodsTo predict oncogenes efficiently, we developed a method, DGAT-onco that analyzed the frequency distribution and functional impacts of mutations in both cancer and natural population. Our method can capture the mutational difference of two population, and provide a comprehensive view of genomics basis underlying cancer development. DGAT-onco was constructed by germline mutations from the 1000 Genomes project and somatic mutations of 33 cancer types from the Cancer Genome Atlas (TCGA) dataset. Its reliability was verified on an independent test set including 19 cancers from other sources. ResultsWe demonstrated that our method is more effective than alternative methods in oncogenes discovering. Using this approach achieves higher classification performance in oncogene discovery than 6 alternative methods, and 22.8% significant genes identified by our method were verified as oncogenes by the Cancer Gene Census (CGC). AvailabilityDGAT-onco is available at https://github.com/zhanghaoyang0/DGAT-onco. Contactyangyd25@mail.sysu.edu.cn or zhaohy8@mail.sysu.edu.cn

bioinformatics↗

bioRxiv2020

PCNA activates the MutLγ endonuclease to promote meiotic crossing over

During meiosis, crossover recombination connects homologous chromosomes to direct their accurate segregation1. Defects in crossing over cause infertility, miscarriage and congenital disease. Accordingly, each pair of chromosomes attains at least one crossover through processes that designate and then implement crossing over with high efficiency2. At the DNA level, crossing over is implemented through the formation and biased resolution of double-Holliday Junction intermediates3-5. A central tenet of crossover resolution is that the two Holliday junctions are resolved in opposite planes by targeting nuclease incisions to specific DNA strands6. Although the endonuclease activity of the MutL{gamma} complex has been implicated in crossover-biased resolution7-12, the mechanisms that activate and direct strand-specific cleavage remain unknown. Here we show that the sliding clamp, PCNA, is important for crossover-biased resolution. In vitro assays with human enzymes show that hPCNA and its loader hRFC are sufficient to activate the hMutL{gamma} endonuclease under physiological conditions. In this context, the hMutL{gamma} endonuclease is further stimulated by a co-dependent activity of the pro-crossover factors hEXO1 and hMutS{gamma}, the latter of which binds Holliday junctions13. hMutL{gamma} also specifically binds a variety of branched DNAs, including Holliday junctions, but canonical resolvase activity is not observed implying that the endonuclease incises adjacent to junction branch points to effect resolution. In vivo, we show that budding yeast RFC facilitates MutL{gamma}-dependent crossing over. Furthermore, PCNA localizes to prospective crossover sites along synapsed chromosomes. These data highlight similarities between crossover-resolution and DNA mismatch repair14-16 and evoke a novel model for crossover-specific dHJ resolution during meiosis.

biochemistry↗

bioRxiv2020

Astroglial FMRP modulates synaptic signaling and behavior phenotypes in FXS mouse model

Fragile X syndrome (FXS) is one of the most common inherited intellectual disability (ID) disorders, in which the loss of FMRP protein induces a range of cellular signaling changes primarily through excess protein synthesis. Although neuron-centered molecular and cellular events underlying FXS have been characterized, how different CNS cell types are involved in typical FXS synaptic signaling changes and behavioral phenotypes is largely unknown. Recent evidence suggests that selective loss of astroglial FMRP is able to dysregulate glutamate uptake, increase spine density, and impair motor-skill learning. Here we investigated the effect of astroglial FMRP on synaptic signaling and FXS-related behavioral and learning phenotypes in astroglial Fmr1 cKO and cON mice in which FMRP expression is selectively diminished or restored in astroglia. We found that selective loss of astroglial FMRP contributes to cortical hyperexcitability by enhancing NMDAR-mediated evoked but not spontaneous miniEPSCs and elongating cortical UP state duration. Selective loss of astroglial FMRP is also sufficient to increase locomotor hyperactivity, significantly diminish social novelty preference, and induce memory acquisition and extinction deficits in astroglial Fmr1 cKO mice. Importantly, re-expression of astroglial FMRP is able to significantly rescue the hyperactivity (evoked NMDAR response, UP state duration, and open field test) and social novelty preference in astroglial Fmr1 cON mice. These results demonstrate a profound role of astroglial FMRP in the evoked synaptic signaling, spontaneously occurring cortical UP states, and FXS-related behavioral and learning phenotypes and provide important new insights in the cell type consideration for the FMRP reactivation strategy.

neuroscience↗

bioRxiv2020

To improve the predictions of binding residues with DNA, RNA, carbohydrate, and peptide via multiple-task deep neural networks

MotivationThe interactions of proteins with DNA, RNA, peptide, and carbohydrate play key roles in various biological processes. The studies of uncharacterized protein-molecules interactions could be aided by accurate predictions of residues that bind with partner molecules. However, the existing methods for predicting binding residues on proteins remain of relatively low accuracies due to the limited number of complex structures in databases. As different types of molecules partially share chemical mechanisms, the predictions for each molecular type should benefit from the binding information with other molecules types. ResultsIn this study, we employed a multiple task deep learning strategy to develop a new sequence-based method for simultaneously predicting binding residues/sites with multiple important molecule types named MTDsite. By combining four training sets for DNA, RNA, peptide, and carbohydrate-binding proteins, our method yielded accurate and robust predictions with AUC values of 0.852, 0836, 0.758, and 0.776 on their respective independent test sets, which are 0.52 to 6.6% better than other state-of-the-art methods. More importantly, this study provides a new strategy to improve predictions by combining multiple similar tasks. Availabilityhttp://biomed.nscc-gz.cn/server/MTDsite/ Contactyangyd25@mail.sysu.edu.cn

bioinformatics↗

bioRxiv2020

Imputing Single-cell RNA-seq data by combining Graph Convolution and Autoencoder Neural Networks

Single-cell RNA sequencing technology promotes the profiling of single-cell transcriptomes at an unprecedented throughput and resolution. However, in scRNA-seq studies, only a low amount of sequenced mRNA in each cell leads to missing detection for a portion of mRNA molecules, i.e. the dropout problem. The dropout event hinders various downstream analysis, such as clustering analysis, differential expression analysis, and inference of gene-to-gene relationships. Therefore, it is necessary to develop robust and effective imputation methods for the increasing scRNA-seq data. In this study, we have developed an imputation method (GraphSCI) to impute the dropout events in scRNA-seq data based on the graph convolution networks. The method takes advantage of low-dimensional representations of similar cells and gene-gene interactions to impute the dropouts. Extensive experiments demonstrated that GraphSCI outperforms other state-of-the-art methods for imputation on both simulated and real scRNA-seq data. Meanwhile, GraphSCI is able to accurately infer gene-to-gene relationships by utilizing the imputed matrix that are concealed by dropout events in raw data.

bioinformatics↗