Biology subjects

Cai, J.

Publications and source records attributed to Cai, J..

At least 127 records · Page 7Linked to original sources

bioRxiv2021

Chromatin-interacting transposon RNAs linking to the core trans-inhibition circuitry for embryonic stem cell identity

Transposable DNA sequences constitute more than half of the human and mouse genomes. A large number of non-coding RNAs, named as transposon RNAs, are derived from these transposable elements. The cis-regulatory function of transposable DNA elements, such as LINE-1 and Alu has been largely explored. But the biological roles of transposon RNAs arent well understood. Here, investigations of RNA-chromatin interactions provide us with comprehensive evidence that specific families of transposon RNAs play roles in trans-regulation linking to the core inhibition circuitry for embryonic stem cell identity. Alternative modes of the RNA-DNA hybrid duplex and protein-recruited RNA scaffold are required for the regulatory activities of transposon RNAs. LINE-1 RNAs co-locating with KAP1 form a negative feedback loop stabilizing the transcription of LINE-1 DNA elements via RNA-DNA hybrids. In another way LINE-1 RNAs, together with the reprogramming three factors and Polycomb repressive complexes, participate in the inhibition on dozens of differentiation-relative genes.

genomics↗

bioRxiv2021

Egocentric Asymmetric Coding in Sensory Cortical Border Cells

Both egocentric and allocentric representations of space are essential to spatial navigation. Although some studies of egocentric coding have been conducted within and around the hippocampal formation, externally anchored egocentric spatial representations have not yet been fully explored. Here we record and identify two subtypes of border cell in the rat primary somatosensory cortex (S1) and secondary visual cortex (V2). Subpopulations of S1 and V2 border cells exhibit rotation-selective asymmetric firing fields in an either clockwise (CW) or counterclockwise (CCW) manner. CW- and CCW-border cells increase their firing rates when animals move unidirectionally along environmental border(s). We demonstrate that both CW- and CCW-border cells fire in an egocentric reference frame relative to environmental borders, maintain preferred directional tunings in rotated, stretched, dark as well as novel arenas, and switch their directional firings in the presence of multi-layer concentric enclosures. These findings may provide rotation-selective egocentric reference frames within a larger spatial navigation system, and point to a common computational principle of spatial coding shared by multiple sensory cortical areas. HighlightsO_LIEgocentric border cells are present in rat S1 and V2 C_LIO_LISubtypes of border cells display egocentric asymmetric coding C_LIO_LIEgocentric and allocentric streams coexist in sensory cortices C_LIO_LIRotation-selective asymmetric firing is robust with environmental manipulations C_LI

neuroscience↗

bioRxiv2021

A compact spatial map in V2 visual cortex

Vision plays a critical role in guiding spatial navigation. A traditional view of the visual cortex is to compute a world-centered map of visual space, and visual neurons exhibit diverse tunings to simple or complex visual features. The neural representation of spatio-visual map in the visual cortex is thought to be transformed from spatial modulation signals at the hippocampal-entorhinal system. Although visual thalamic and cortical neurons have been shown to be modulated by spatial signals during navigation, the exact source of spatially modulated neurons within the visual circuit has never been identified, and the neural correlate underpinning a visuospatial or spatio-visual map remains elusive. To search for direct visuospatial and visuodirectional signals, here we record in vivo extracellular spiking activity in the secondary visual cortex (V2) from freely foraging rats in a naturalistic environment. We identify that V2 neurons forms a complete spatio-visual map with a wide range of spatial tunings, which resembles the classical spatial map that includes the place, head-direction, border, grid and conjunctive cells reported in the hippocampal-entorhinal network. These spatially tuned V2 neurons display stable responses to external visual cues, and are robust with respect to non- spatial environmental changes. Spatially and directionally tuned V2 neuronal firing persists in darkness, suggesting that this spatio-visual map is not completely dependent on visual inputs. Identification of functionally distinct spatial cell types in visual cortex expands its classical role of information coding beyond a retinotopic map of the eye-centered world.

neuroscience↗

bioRxiv2021

Environment-Dependent Switch between Two Immune Systems against Ralstonia solanacearum Infections in Pepper

O_LIAlthough plant diseases generally cause more severe symptoms under conditions of high temperature and high humidity(HTHH), how plant respond to pathogen attack under this condition remains elusive. C_LIO_LIAs an example, we herein comparatively studied pepper(Capsicum annuum) immunity against Ralstonia solanacearum under HTHH and ambient temperature by approaches of reverse genetics. We found that pepper respond to R.solanacearum infection by activating salicylic-acid- and jasmonic-acid-mediated immunity at ambient temperature. Under HTHH, However, it no longer activates JA-mediated immunity and activates only transient SA signaling at the early stage of R.solanacearum infection, but instead induces cytokinin mediated immunity. C_LIO_LIThese two immune systems are positively regulated by CaWRKY40 via binding the WT-box and a novel W-box like (WL) box, respectively, in an environment-dependent manner: CaWRKY40 is activated upon R. solanacearum infection under HTHH, thereby upregulating ISOPENTENYLTRANSFERASE5(IPT5). The resulting cytokinin then works synergistically with CaWRKY40 in activating a subset of glutathione S-transferase genes via chromatin activation and WL-box binding, but prevent CaWRKY40 from activating ISOCHORISMATE SYNTHASE1 (ICS1) or SA-/JA-dependent pathogenesis-related genes by chromatin inactivation or by blocking WT-box binding. C_LIO_LIThese results highlight the specific pepper immune response to R. solanacearum infection under HTHH and its synergistic activation by CaWRKY40 and cytokinins. C_LI

plant biology↗

bioRxiv2020

Integrating thermodynamic and enzymatic constraints into genome-scale metabolic models

Stoichiometric genome-scale metabolic network models (GEMs) have been widely used to predict metabolic phenotypes. In addition to stoichiometric ratios, other constraints such as enzyme availability and thermodynamic feasibility can also limit the phenotype solution space. Extended GEM models considering either enzymatic or thermodynamic constraints have been shown to improve prediction accuracy. In this paper, we propose a novel method that integrates both enzymatic and thermodynamic constraints in a single Pyomo modeling framework (ETGEMs). We applied this method to construct the EcoETM, the E. coli metabolic model iML1515 with enzymatic and thermodynamic constraints. Using this model, we calculated the optimal pathways for cellular growth and the production of 22 metabolites. When comparing the results with those of iML1515 and models with one of the two constraints, we observed that many thermodynamically unfavorable and/or high enzyme cost pathways were excluded from EcoETM. For example, the synthesis pathway of carbamoyl-phosphate (Cbp) from iML1515 is both thermodynamically unfavorable and enzymatically costly. After introducing the new constraints, the production pathways and yields of several Cbp-derived products (e.g. L-arginine, orotate) calculated using EcoETM were more realistic. The results of this study demonstrate the great application potential of metabolic models with multiple constraints for pathway analysis and phenotype predication.

bioinformatics↗

bioRxiv2020

scCapsNet-mask: an automatic version of scCapsNet

SummaryRecently we developed scCapsNet, an interpretable deep learning cell type classifier for single cell RNA sequencing data, based on capsule network. However, the running process of scCapsNet is not fully automatic, in which a manual intervention is required for getting the final results. Here we present scCapsNet-mask, an updated version of scCapsNet that utilizes a mask to fully automate the running process of scCapsNet. scCapsNet-mask could constrain the internal parameter coupling coefficients and result in a one to one correspondence between the primary capsule and type capsule. Based on those bijective mapping between primary capsule and type capsule, the model could automatically extract the cell type related genes according to weight matrix connecting input and primary capsule, without a need for manual inspection of the relationship between primary capsules and type capsules. The scCapsNet-mask is evaluated on two single cell RNA sequence datasets. The results show that scCapsNet-mask not only retains the merits of the original scCapsNet with high classification accuracy and high interpretability, but also has the virtue of automatic processing.

bioinformatics↗

bioRxiv2020

Differences in the gut microbiomes of distinct ethnicities within the same geographic area are linked to host metabolic health

BackgroundThe human gut microbiota exhibits marked variation around the world, which has been attributed to dietary intake and other environmental factors. However, the degree to which ethnicity-associated differences in gut microbial community structure and function are maintained following immigration or in the context of metabolic disease is poorly understood. ResultsWe conducted a multi-omic study of 46 lean and obese East Asian and White participants living in the San Francisco Bay Area. 16S rRNA gene sequencing revealed significant differences between ethnic groups in bacterial richness and community structure. White individuals were enriched for the mucin-degrading Akkermansia muciniphila. East Asian participants had increased levels of multiple bacterial phyla, fermentative pathways detected by metagenomics, and the short-chain fatty acid end products acetate, propionate, and isobutyrate. Differences in the gut microbiota between the East Asian and White groups could not be explained by reported dietary intake, were more pronounced in lean individuals, and were associated with current geographical location. Microbiome transplantations into germ-free mice confirmed that the differences in the gut microbiota of the East Asian and White individuals we analyzed are independent of diet and that they differentially impact host body weight and adiposity in genetically identical mouse recipients. ConclusionsThe reported findings emphasize the utility of studying diverse ethnic groups within a defined geographical location and provide a starting point for dissecting the mechanisms contributing to the complex interactions between the gut microbiome and ethnicity-associated lifestyle, demographic, metabolic, and genetic factors.

microbiology↗

bioRxiv2020

VirusDIP: Virus Data Integration Platform

MotivationThe Coronavirus Disease 2019 (COVID-19) pandemic poses a huge threat to human public health. Viral sequence data plays an important role in the scientific prevention and control of epidemics. A comprehensive virus database will be vital useful for virus data retrieval and deep analysis. To promote sharing of virus data, several virus databases and related analyzing tools have been created. ResultsTo facilitate virus research and promote the global sharing of virus data, we present here VirusDIP, a one-stop service platform for archive, integration, access, analysis of virus data. It accepts the submission of viral sequence data from all over the world and currently integrates data resources from the National GeneBank Database (CNGBdb), Global initiative on sharing all influenza data (GISAID), and National Center for Biotechnology Information (NCBI). Moreover, based on the comprehensive data resources, BLAST sequence alignment tool and multi-party security computing tools are deployed for multi-sequence alignment, phylogenetic tree building and global trusted sharing. VirusDIP is gradually establishing cooperation with more databases, and paving the way for the analysis of virus origin and evolution. All public data in VirusDIP are freely available for all researchers worldwide. Availabilityhttps://db.cngb.org/virus/ Contactweixiaofeng@cngb.org

bioinformatics↗

bioRxiv2020

CNSA: a data repository for archiving omics data

With the application and development of high-throughput sequencing technology in life and health sciences, massive multi-dimensional biological data brings the problem of efficient management and utilization. Database development and biocuration are the prerequisites for the reuse of these big data. Here, relying on China National GeneBank (CNGB), we present CNGB Sequence Archive (CNSA) for archiving omics data, including raw sequencing data and its analytical data and related metadata which are organized into six objects, namely Project, Sample, Experiment, Run, Assembly, and Variation at present. Moreover, CNSA has created the correlation model of living samples, sample information, and analytical data on some projects, so that all data can be traced throughout the life cycle from the living sample to the sample information to the analytical data. Complying with the data standards commonly used in the life sciences, CNSA is committed to building a comprehensive and curated data repository for the storage, management and sharing of omics data, improving the data standards, and providing free access to open data resources for worldwide scientific communities to support academic research and the bio-industry. Database URL: https://db.cngb.org/cnsa/

genomics↗

bioRxiv2020

A long noncoding RNA, LOC157273, is the effector transcript at the chromosome 8p23.1-PPP1R3B metabolic traits and type 2 diabetes risk locus

AimsCausal transcripts at genomic loci associated with type 2 diabetes are mostly unknown. The chr8p23.1 variant rs4841132, associated with an insulin resistant diabetes risk phenotype, lies in the second exon of a long non-coding RNA (lncRNA) gene, LOC157273, located 175 kilobases from PPP1R3B, which encodes a key protein regulating insulin-mediated hepatic glycogen storage in humans. We hypothesized that LOC157273 regulates expression of PPP1R3B in human hepatocytes. MethodsWe tested our hypothesis using Stellaris fluorescent in-situ hybridization to assess subcellular localization of LOC157273; siRNA knockdown of LOC157273, followed by RT-PCR to quantify LOC157273 and PPP1R3B expression; RNA-seq to quantify the whole-transcriptome gene expression response to LOC157273 knockdown and an insulin-stimulated assay to measure hepatocyte glycogen deposition before and after knockdown. ResultsWe found that siRNA knockdown decreased LOC157273 transcript levels by approximately 80%, increased PPP1R3B mRNA levels by 1.7-fold and increased glycogen deposition by >50% in primary human hepatocytes. An A/G heterozygous carrier (vs. three G/G carriers) had reduced LOC157273 abundance due to reduced transcription of the A allele and increased PPP1R3B expression and glycogen deposition. ConclusionWe show that the lncRNA LOC157273 is a negative regulator of PPP1R3B expression and glycogen deposition in human hepatocytes and the causal transcript at an insulin resistant type 2 diabetes risk locus.

genomics↗

bioRxiv2020

Multiple approaches for massively parallel sequencing of HCoV-19 genomes directly from clinical samples

COVID-19 has caused a major epidemic worldwide, however, much is yet to be known about the epidemiology and evolution of the virus. One reason is that the challenges underneath sequencing HCoV-19 directly from clinical samples have not been completely tackled. Here we illustrate the application of amplicon and hybrid capture (capture)-based sequencing, as well as ultra-high-throughput metatranscriptomic (meta) sequencing in retrieving complete genomes, inter-individual and intra-individual variations of HCoV-19 from clinical samples covering a range of sample types and viral load. We also examine and compare the bias, sensitivity, accuracy, and other characteristics of these approaches in a comprehensive manner. This is, to date, the first work systematically implements amplicon and capture approaches in sequencing HCoV-19, as well as the first comparative study across methods. Our work offers practical solutions for genome sequencing and analyses of HCoV-19 and other emerging viruses.

genomics↗

bioRxiv2020

Aerodynamic Characteristics and RNA Concentration of SARS-CoV-2 Aerosol in Wuhan Hospitals during COVID-19 Outbreak

BackgroundThe ongoing outbreak of COVID-19 has spread rapidly and sparked global concern. While the transmission of SARS-CoV-2 through human respiratory droplets and contact with infected persons is clear, the aerosol transmission of SARS-CoV-2 has been little studied. MethodsThirty-five aerosol samples of three different types (total suspended particle, size segregated and deposition aerosol) were collected in Patient Areas (PAA) and Medical Staff Areas (MSA) of Renmin Hospital of Wuhan University (Renmin) and Wuchang Fangcang Field Hospital (Fangcang), and Public Areas (PUA) in Wuhan, China during COVID-19 outbreak. A robust droplet digital polymerase chain reaction (ddPCR) method was employed to quantitate the viral SARS-CoV-2 RNA genome and determine aerosol RNA concentration. ResultsThe ICU, CCU and general patient rooms inside Renmin, patient hall inside Fangcang had undetectable or low airborne SARS-CoV-2 concentration but deposition samples inside ICU and air sample in Fangcang patient toilet tested positive. The airborne SARS-CoV-2 in Fangcang MSA had bimodal distribution with higher concentration than those in Renmin during the outbreak but turned negative after patients number reduced and rigorous sanitization implemented. PUA had undetectable airborne SARS-CoV-2 concentration but obviously increased with accumulating crowd flow. ConclusionsRoom ventilation, open space, proper use and disinfection of toilet can effectively limit aerosol transmission of SARS-CoV-2. Gathering of crowds with asymptomatic carriers is a potential source of airborne SARS-CoV-2. The virus aerosol deposition on protective apparel or floor surface and their subsequent resuspension is a potential transmission pathway and effective sanitization is critical in minimizing aerosol transmission of SARS-CoV-2.

microbiology↗

bioRxiv2020

Specific ACE2 Expression in Cholangiocytes May Cause Liver Damage After 2019-nCoV Infection

A newly identified coronavirus, 2019-nCoV, has been posing significant threats to public health since December 2019. ACE2, the host cell receptor for severe acute respiratory syndrome coronavirus (SARS), has recently been demonstrated in mediating 2019-nCoV infection. Interestingly, besides the respiratory system, substantial proportion of SARS and 2019-nCoV patients showed signs of various degrees of liver damage, the mechanism and implication of which have not yet been determined. Here, we performed an unbiased evaluation of cell type specific expression of ACE2 in healthy liver tissues using single cell RNA-seq data of two independent cohorts, and identified specific expression in cholangiocytes. The results indicated that virus might directly bind to ACE2 positive cholangiocytes but not necessarily hepatocytes. This finding suggested the liver abnormalities of SARS and 2019-nCoV patients may not be due to hepatocyte damage, but cholangiocyte dysfunction and other causes such as drug induced and systemic inflammatory response induced liver injury. Our findings indicate that special care of liver dysfunction should be installed in treating 2019-nCoV patients during the hospitalization and shortly after cure.

genomics↗

bioRxiv2020

Structural insights into the insecticidal Vip3A toxin of Bacillus thuringiensis

The vegetative insecticidal proteins (Vips) secreted by Bacillus thuringiensis are regarded as the new generation of insecticidal toxins because they have different insecticidal properties compared with commonly applied insecticidal crystal proteins (Cry toxins). Vip3A toxin, representing the vast majority of Vips, has been used commercially in transgenic crops and bio-insecticides. However, the lack of both structural information of Vip3A and a clear understanding of its insecticidal mechanism at the molecular level, limits its further development and broader application. Here we present the first crystal structure of the Vip3A toxin in an activated form. Since all members of this insecticidal protein family are highly conserved, the structure of Vip3A provides unique insight into the general domain architecture and protein fold of the Vip3 family of insecticidal toxins. Our structural analysis reveals a four-domain organization, featuring a potential membrane insertion region, a receptor binding domain, and two glycan binding domains of activated Vip3A. We further identify the specific glycan moieties recognized by Vip3A through a glycan array screen. Taken together, these findings provide insights into the mode of action of Vip3 family of insecticidal toxins, and will boost the development of Vip3 into more efficient bio-insecticides.

microbiology↗

bioRxiv2020

Alteration of fecal microbial compositions and bacterial taxa in female osteoporotic patients

BackgroundGut microbiota, mainly characterized by fecal bacterial compositions, affects human immune system and pathophysiological development. Our aim was to measure the quantitative differences of fecal bacterial compositions between osteoporotic patients and healthy subjects, and to identify novel bacterial taxa that speculate the incidence of osteoporosis in female. MethodWe recruited 104 female subjects, including 45 osteoporotic individuals and 59 healthy control. Fecal samples were collected for further analysis by 16S rRNA quantitative arrays and bioinformatics analysis. ResultsAnalyses of - and {beta}-diversity demonstrated that the diversity and composition of fecal bacterial compositions were both significant different in osteoporosis group, as compared with healthy group. Multiple bacterial genera were significantly increased (e.g., Roseburia and Bacteroides) or decreased (e.g., Streptococcus and Dorea) in the osteoporotic cases. Furthermore, the osteoporosiscould be efficiently determined by the random forest model based on differential taxa (area under ROC curve = 0.93). ConclusionThere were obvious different fecal microbial characteristics between female osteoporosis and healthy subjects. These findings provided evidence for understanding the host-gut microbiota interplay in female osteoporosis, and supported clinical applications of gut microbiota analysis for female osteoporosis diagnosis

microbiology↗

bioRxiv2020

Ursodeoxycholic acid (UDCA) mitigates the host inflammatory response during Clostridioides difficile infection by altering gut bile acids which attenuates NF-κB signaling via bile acid activated receptors

Clostridioides difficile infection (CDI) is associated with increasing morbidity and mortality posing an urgent threat to public health. Recurrence of CDI after successful treatment with antibiotics is high, thus necessitating discovery of novel therapeutics against this enteric pathogen. Administration of the secondary bile acid ursodeoxycholic acid (UDCA, ursodiol) inhibits the life cycle of various strains of C. difficile in vitro, suggesting the FDA approved formulation of UDCA, known as ursodiol, may be able to restore colonization resistance against C. difficile in vivo. However, the mechanism(s) by which ursodiol is able to restore colonization resistance against C. difficile remains unknown. Here, we confirmed that ursodiol inhibits C. difficile R20291 spore germination and outgrowth, growth, and toxin activity in a dose dependent manner in vitro. In a murine model of CDI, exogenous administration of ursodiol resulted in significant alterations in the bile acid metabolome with little to no changes in gut microbial community structure. Ursodiol pretreatment resulted in attenuation of CDI pathogenesis early in the course of disease, which coincided with alterations in the cecal and colonic inflammatory transcriptome, bile acid activated receptors nuclear farnesoid X receptor (FXR), and transmembrane G protein-coupled membrane receptor 5 (TGR5), which are able to modulate the innate immune response through signaling pathways such as NF-{kappa}B. Although ursodiol pretreatment did not result in a consistent decrease in the C. difficile life cycle in vivo, it was able to attenuate an overly robust inflammatory response that is detrimental to the host during CDI. Ursodiol remains a viable non-antibiotic treatment and/or prevention strategy against CDI. Likewise, modulation of the host innate immune response via bile acid activated receptors, FXR and TGR5, represents a new potential treatment strategy for patients with CDI. ImportanceThe clinical utility of ursodiol for prevention of recurrent CDI is currently in Phase 4 clinical trials. However, the mechanism by which ursodiol exerts its impacts on C. difficile pathogenesis is poorly understood. Herein, we demonstrated that ursodiol pretreatment attenuates CDI pathogenesis early in the course of disease in mice, which coincides with alterations in the cecal and colonic inflammatory transcriptome, bile acid activated receptors nuclear farnesoid X receptor (FXR), and transmembrane G protein-coupled membrane receptor 5 (TGR5), which are able to modulate the innate immune response through signaling pathways such as NF-{kappa}B. Ursodiol attenuated an overly robust inflammatory response that is detrimental to the host during CDI, and thus remains a viable non-antibiotic treatment and/or prevention strategy against CDI. Likewise, modulation of the host innate immune response via bile acid activated receptors, FXR and TGR5, represents a new potential treatment strategy for patients with CDI. AbbreviationsMCA - -Muricholic acid; {beta}MCA -{beta}-Muricholic acid; {omega}MCA -{omega}-Muricholic acid; CA - Cholic acid; CDCA - Chenodeoxycholic acid; DCA - Deoxycholic acid; GCDCA - Glycochenodeoxycholic acid; GDCA - Glycodeoxycholic acid; GLCA - Glycolithocholic acid; GUDCA - Glycoursodeoxycholic acid; HCA - Hyodeoxycholic acid; iDCA - Isodeoxycholic acid; iLCA - Isolithocholic acid; LCA - Lithocholic acid; TCA - Taurocholic acid; TCDCA - Taurochenodeoxycholic acid; TDCA - Taurodeoxycholic acid; THCA - Taurohyodeoxycholic acid; TUDCA - Tauroursodeoxycholic acid; T{beta}MCA- Tauro-{beta}-muricholic acid; T{omega}MCA -Tauro {omega}-muricholic acid; UDCA Ursodeoxycholic acid.

microbiology↗

bioRxiv2019

The virus integrations in gut microbes associated with dysbiosis of microbial community in tumorigenesis of colorectal carcinoma

Patients with colorectal cancer (CRC) have a different gut microbial and viral communities from healthy individuals. But little is known about the ways and functions of interaction of virus-bacteria, let alone its correlation with the aetiology of CRC. In this study we aimed to identify the association between the genetic integration of virusbacteria and the expansion of some microbial population during tumorigenesis of human colorectum. Using a gut metagenomics sequencing data of healthy controls, advanced adenoma and carcinoma patients, to our knowledge, we demonstrate for the first time that the viral genetic integrations in gut microbes tend to occur in CRC patients and are potentially associated with the carcinogenesis. We found that almost all of the genetic integrations were happened between bacteriophages and bacteria, which could be influenced by the abundance of the phage communities. Importantly, the integrations of phage-carried positive effective genes offered selective advantages to the commensal and potential pathogenic bacteria, as a result, potentially led to a microbial dysbiosis along with the increasing bacterial diversity in carcinoma patients. Consequently, our work opens a new way to understand the carcinogenicity in complex intestinal ecosystems.

microbiology↗

bioRxiv2019

Neural Gene Network Constructor: A Neural Based Model for Reconstructing Gene Regulatory Network

Reconstructing gene regulatory networks (GRNs) and inferring the gene dynamics are important to understand the behavior and the fate of the normal and abnormal cells. Gene regulatory networks could be reconstructed by experimental methods or from gene expression data. Recent advances in Single Cell RNA sequencing technology and the computational method to reconstruct trajectory have generated huge scRNA-seq data tagged with additional time labels. Here, we present a deep learning model "Neural Gene Network Constructor" (NGNC), for inferring gene regulatory network and reconstructing the gene dynamics simultaneously from time series gene expression data. NGNC is a model-free heterogenous model, which can reconstruct any network structure and non-linear dynamics. It consists of two parts: a network generator which incorporating gumbel softmax technique to generate candidate network structure, and a dynamics learner which adopting multiple feedforward neural networks to predict the dynamics. We compare our model with other well-known frameworks on the data set generated by GeneNetWeaver, and achieve the state of the arts results both on network reconstruction and dynamics learning.

bioinformatics↗