Variational graph auto-encoders for miRNA-disease association prediction.

Cumulative experimental studies have demonstrated the critical roles of microRNAs (miRNAs) in the diverse fundamental and important biological processes, and in the development of numerous complex human diseases. Thus, exploring the relationships between miRNAs and diseases is helpful with understanding the mechanisms, the detection, diagnosis, and treatment of complex diseases. As the identification of miRNA-disease associations via traditional biological experiments is time-consuming and expensive, an effective computational prediction method is appealing. In this study, we present a deep learning framework with variational graph auto-encoder for miRNA-disease association prediction (VGAE-MDA). VGAE-MDA first gets the representations of miRNAs and diseases from the heterogeneous networks constructed by miRNA-miRNA similarity, disease-disease similarity, and known miRNA-disease associations. Then, VGAE-MDA constructs two sub-networks: miRNA-based network and disease-based network. Combining the representations based on the heterogeneous network, two variational graph auto-encoders (VGAE) are deployed for calculating the miRNA-disease association scores from two sub-networks, respectively. Lastly, VGAE-MDA obtains the final predicted association score for a miRNA-disease pair by integrating the scores from these two trained networks. Unlike the previous model, the VGAE-MDA can mitigate the effect of noises from random selection of negative samples. Besides, the use of graph convolutional neural (GCN) network can naturally incorporate the node features from the graph structure while the variational autoencoder (VAE) makes use of latent variables to predict associations from the perspective of data distribution. The experimental results show that VGAE-MDA outperforms the state-of-the-art approaches in miRNA-disease association prediction. Besides, the effectiveness of our model has been further demonstrated by case studies.

Ensemble disease gene prediction by clinical sample-based networks.

BACKGROUND: Disease gene prediction is a critical and challenging task. Many computational methods have been developed to predict disease genes, which can reduce the money and time used in the experimental validation. Since proteins (products of genes) usually work together to achieve a specific function, biomolecular networks, such as the protein-protein interaction (PPI) network and gene co-expression networks, are widely used to predict disease genes by analyzing the relationships between known disease genes and other genes in the networks. However, existing methods commonly use a universal static PPI network, which ignore the fact that PPIs are dynamic, and PPIs in various patients should also be different. RESULTS: To address these issues, we develop an ensemble algorithm to predict disease genes from clinical sample-based networks (EdgCSN). The algorithm first constructs single sample-based networks for each case sample of the disease under study. Then, these single sample-based networks are merged to several fused networks based on the clustering results of the samples. After that, logistic models are trained with centrality features extracted from the fused networks, and an ensemble strategy is used to predict the finial probability of each gene being disease-associated. EdgCSN is evaluated on breast cancer (BC), thyroid cancer (TC) and Alzheimer's disease (AD) and obtains AUC values of 0.970, 0.971 and 0.966, respectively, which are much better than the competing algorithms. Subsequent de novo validations also demonstrate the ability of EdgCSN in predicting new disease genes. CONCLUSIONS: In this study, we propose EdgCSN, which is an ensemble learning algorithm for predicting disease genes with models trained by centrality features extracted from clinical sample-based networks. Results of the leave-one-out cross validation show that our EdgCSN performs much better than the competing algorithms in predicting BC-associated, TC-associated and AD-associated genes. de novo validations also show that EdgCSN is valuable for identifying new disease genes.

Enhancing the prediction of disease-gene associations with multimodal deep learning.

MOTIVATION: Computationally predicting disease genes helps scientists optimize the in-depth experimental validation and accelerates the identification of real disease-associated genes. Modern high-throughput technologies have generated a vast amount of omics data, and integrating them is expected to improve the accuracy of computational prediction. As an integrative model, multimodal deep belief net (DBN) can capture cross-modality features from heterogeneous datasets to model a complex system. Studies have shown its power in image classification and tumor subtype prediction. However, multimodal DBN has not been used in predicting disease-gene associations. RESULTS: In this study, we propose a method to predict disease-gene associations by multimodal DBN (dgMDL). Specifically, latent representations of protein-protein interaction networks and gene ontology terms are first learned by two DBNs independently. Then, a joint DBN is used to learn cross-modality representations from the two sub-models by taking the concatenation of their obtained latent representations as the multimodal input. Finally, disease-gene associations are predicted with the learned cross-modality representations. The proposed method is compared with two state-of-the-art algorithms in terms of 5-fold cross-validation on a set of curated disease-gene associations. dgMDL achieves an AUC of 0.969 which is superior to the competing algorithms. Further analysis of the top-10 unknown disease-gene pairs also demonstrates the ability of dgMDL in predicting new disease-gene associations. AVAILABILITY AND IMPLEMENTATION: Prediction results and a reference implementation of dgMDL in Python is available on https://github.com/luoping1004/dgMDL. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

[Chemical constituents of ethyl acetate fraction from seed melon (Citrulluslanatus ssp. vulgaris var. megalaspermus)].

In this paper, the chemical composition of ethyl acetate parts of seed melon were studied by using ethanol re-flux method, extraction method, and isolated by column chromatography oversilica gel and Sephadex LH-20 and HPLC. The structures of the separated compounds were identified by physical-chemical methods and spectral data such as MS, ¹H-NMR, ¹³C-NMR, etc. 12 compounds were got from the plant including one new compound, 4-hydroxymet-hyl-2-methoxyphenyl 1-O-ß-D-[6'-O-(4″-hydroxybenzoyl)-glucopyranoside] (1) and 11 known compounds, uracil (2), thymine (3), 2'-deoxyuridine (4), 7,8-dimethylalloxazine (5), indole-3-carboxylic acid (6), ß-adenosine (7), 4-hydroxybenzoic acid (8), p-coumaric acid (9), cucumegastigmanesⅠ (10), 3'-methoxyl-quercetin-7-O-ß-D-glucopyranoside (11) and 3,3'-dimethyloxy-4,4'-dihydroxy-9,9'-monoepoxy lignan (12).

Nonlinear-model-based analysis methods for time-course gene expression data.

Microarray technology has produced a huge body of time-course gene expression data and will continue to produce more. Such gene expression data has been proved useful in genomic disease diagnosis and drug design. The challenge is how to uncover useful information from such data by proper analysis methods such as significance analysis and clustering analysis. Many statistic-based significance analysis methods and distance/correlation-based clustering analysis methods have been applied to time-course expression data. However, these techniques are unable to account for the dynamics of such data. It is the dynamics that characterizes such data and that should be considered in analysis of such data. In this paper, we employ a nonlinear model to analyse time-course gene expression data. We firstly develop an efficient method for estimating the parameters in the nonlinear model. Then we utilize this model to perform the significance analysis of individually differentially expressed genes and clustering analysis of a set of gene expression profiles. The verification with two synthetic datasets shows that our developed significance analysis method and cluster analysis method outperform some existing methods. The application to one real-life biological dataset illustrates that the analysis results of our developed methods are in agreement with the existing results.

Nonlinear model-based method for clustering periodically expressed genes.

Clustering periodically expressed genes from their time-course expression data could help understand the molecular mechanism of those biological processes. In this paper, we propose a nonlinear model-based clustering method for periodically expressed gene profiles. As periodically expressed genes are associated with periodic biological processes, the proposed method naturally assumes that a periodically expressed gene dataset is generated by a number of periodical processes. Each periodical process is modelled by a linear combination of trigonometric sine and cosine functions in time plus a Gaussian noise term. A two stage method is proposed to estimate the model parameter, and a relocation-iteration algorithm is employed to assign each gene to an appropriate cluster. A bootstrapping method and an average adjusted Rand index (AARI) are employed to measure the quality of clustering. One synthetic dataset and two biological datasets were employed to evaluate the performance of the proposed method. The results show that our method allows the better quality clustering than other clustering methods (e.g., k-means) for periodically expressed gene data, and thus it is an effective cluster analysis method for periodically expressed gene data.

[Dynamic QTL and correlated characters of tomato soluble solid content.].

Two hundred and six F2:3 families from the cross between TD22 and HT-1-1-1-1 were used for dynamic QTL research of tomato soluble solid content and correlative traits, and correlation analysis of soluble solid content (SSC) with fruit weight (FW), fruit shape index (FSI), soluble sugar, vitamin C (VC), and organic acid at three different development stages. The results showed that there were differences in QTL loci for soluble solid content during the three stages of tomato fruit development. Four and eight QTLs were detected in green ripe stage and red ripe stage, respectively. These QTLs showed dynamic changes, and two markers LEaat006 and Tomato|TC162363 were detected in two stages, which might be useful in molecular-marker assisted selection (MAS). The result also showed that there was extremely significant difference in SSC at the three different stages, and its main correlative traits were different at different stages. Soluble solid content was positively correlated with soluble sugar, but negatively correlated with FW at green ripe stage; SSC was positively correlated with soluble sugar and organic acid at yellow ripe stage; SSC was positively correlated with soluble sugar and organic acid, but negatively correlated with fruit weight at red ripe stage. Based on correlation analysis of these traits, linear regression model was constructed. Non-tested varieties were used to test the fitness, and the result showed that it is well fitted, and the fitness is above 95%.

Two new phenolic glycosides from the seeds of Citrullus lanatus.

Two new phenolic glycosides, named lanatusosides C (1) and D (2), together with four known compounds (3-6), were isolated from the seeds of Citrullus lanatus. Among them, compounds 3 and 4 were isolated from Cucurbitaceae for the first time, and compound 5 was reported from this plant for the first time. Their structures were elucidated by means of extensive spectral analysis, including HR-ESI-MS, 1H and 13C NMR techniques. The isolated new compounds were evaluated for cytotoxic activity against HepG2 cell line, of which compound 1 demonstrated weak cytotoxicity against the tested cell line.

[The early diagnosis and clinical analysis of 57 cases of acquired hemophagocytic lymphohistiocytosis].

OBJECTIVE: To explore the significance of NK cell activity, interleukin-2 receptors (sCD(25)) and glycosylated ferritin in the early diagnostic of acquired hemophagocytic lymphohistiocytosis (HLH). METHODS: 57 patients suspected of HLH from June 2005 to May 2008 and 25 healthy subjects were enrolled in the study. The patients suspected of HLH were divided into three groups i.e. (1) a group with diagnosis confirmed at first visit; (2) a group with diagnosis confirmed at subsequent visit and (3) a group with diagnosis unconfirmed according to HLH-2004 diagnostic criteria. Healthy subjects were enrolled as control. NK cell activity was determined with a released LDH assay. The percentage of glycosylated ferritin was determined with phytohemagglutinin adsorption assay. sCD(25) was examined with ELISA double antibody sandwich assay. We compared the coincidence of each diagnostic index before and after diagnosis. RESULTS: The median percentage of NK cell activity was significantly lower in the first group (18.3 +/- 5.6)% and the second (16.7 +/- 6.7)% than that in the third group (33.4 +/- 6.8)% or in the controls (36.6 +/- 5.0)%. The median percentage of glycosylated ferritin was also significantly lower in the first group (15.4 +/- 2.0)% and the second group (16.9 +/- 3.4)% than that in the third group (40.4 +/- 3.0)% or in the controls (45.2 +/- 2.2)%. Meanwhile, the median level of sCD(25) was significantly higher in the first group (12 916 +/- 4328) ng/L and the second group (12 117 +/- 5465) ng/L than that in the third group (4728 +/- 1482) ng/L or in the controls (3841 +/- 993) ng/L. Furthermore, NK cell activity, sCD(25) and glycosylated ferritin were abnormal in all the patients in the early stage of HLH. CONCLUSION: NK cell activity, sCD(25) and glycosylated ferritin may be helpful markers for the early diagnosis of HLH.

Disease Gene Prediction by Integrating PPI Networks, Clinical RNA-Seq Data and OMIM Data.

Disease gene prediction is a challenging task that has a variety of applications such as early diagnosis and drug development. The existing machine learning methods suffer from the imbalanced sample issue because the number of known disease genes (positive samples) is much less than that of unknown genes which are typically considered to be negative samples. In addition, most methods have not utilized clinical data from patients with a specific disease to predict disease genes. In this study, we propose a disease gene prediction algorithm (called dgSeq) by combining protein-protein interaction (PPI) network, clinical RNA-Seq data, and Online Mendelian Inheritance in Man (OMIN) data. Our dgSeq constructs differential networks based on rewiring information calculated from clinical RNA-Seq data. To select balanced sets of non-disease genes (negative samples), a disease-gene network is also constructed from OMIM data. After features are extracted from the PPI networks and differential networks, the logistic regression classifiers are trained. Our dgSeq obtains AUC values of 0.88, 0.83, and 0.80 for identifying breast cancer genes, thyroid cancer genes, and Alzheimer's disease genes, respectively, which indicates its superiority to other three competing methods. Both gene set enrichment analysis and predicted results demonstrate that dgSeq can effectively predict new disease genes.

Circulating T lymphocyte subsets, cytokines, and immune checkpoint inhibitors in patients with bipolar II or major depression: a preliminary study.

This study aimed to investigate the less known activation pattern of T lymphocyte populations and immune checkpoint inhibitors on immunocytes in patients with bipolar II disorder depression (BD) or major depression (MD). A total of 23 patients with BD, 22 patients with MD, and 20 healthy controls (HCs) were recruited. The blood cell count of T lymphocyte subsets and the plasma level of cytokines (IL-2, IL-4, IL-6, IL-10, TNF-α, and IFN-γ) were selectively investigated. The expression of T-cell immunoglobulin and mucin-domain containing-3 (TIM-3), programmed cell death protein 1 (PD-1) and its ligands, PD-L1 and PD-L2, on T lymphocytes and monocytes, was detected. In results, blood proportion of cytotoxic T cells significantly decreased in BD patients than in either MD patients or HCs. The plasma level of IL-6 increased in patients with BD and MD. The expression of TIM-3 on cytotoxic T cells significantly increased, whereas the expression of PD-L2 on monocytes significantly decreased in patients with BD than in HCs. These findings extended our knowledge of the immune dysfunction in patients with affective disorders.

Protective effects of seed melon extract on CCl4-induced hepatic fibrosis in mice.

ETHNOPHARMACOLOGICAL RELEVANCE: Citrullus lanatus ssp. vulgaris var. megalaspermus Lin et Chao, was also known as watermelon belongs to family Cucurbitaceae, variously used as healthy food and in the treatment of liver and lungs problems. Currently, Citrullus lanatus has become a major economic crop of medicinal and edible effects with regional characteristics. AIM: This study was designed to evaluate the hepatoprotective and antioxidant activity of the seed melon (Citrullus lanatus ssp. vulgaris var. megalaspermus Lin et Chao) extract (SME) against carbon tetrachloride (CCl4) induced hepatic fibrosis in mice. MATERIALS AND METHODS: In this study, mice were randomly divided into 7 groups, including normal control, model, silymarin tablets as the positive control, SME 100, 200, 400, and 800mg/kg. After 8 weeks, activities of serum alanine aminotransferase (ALT), aspartate aminotransferase (AST), triglycerides (TG), hyaluronic acid (HA) and laminin (LN) were checked. The levels of antioxidant enzymes such as superoxide dismutase (SOD), glutataion (GSH) and glutathione peroxidase (GSH-Px) were determined after SME administration. The hydroxyproline (HYP) levels, malondialdehyde (MDA) levels and histopathologic examinations of hepatocyte fibrosis were also determined. Additionally, effects of SME on alpha-smooth muscle actin (α-SMA) and transforming growth factor beta-1(TGF-ß1) protein expressions were determined. RESULTS: We found that SME could significantly lower the serum levels of hepatic enzyme markers AST, ALT, HA and LN (P<0.01). Compared with the CCl4-only treatment group, levels of hepatic SOD and GSH-Px were significantly increased, and the MDA levels were remarkably decreased in mice treated by SME at medium dose (400mg/kg) and high dose (800mg/kg) (P<0.01). A histological examination of the liver showed that lesions, including necrosis, lymphocyte infiltration and fatty degeneration, were partially healed by treatment with SME. The results of protein expressions studies displayed that SME could inhibit α-SMA and TGF-ß1 protein expression (P<0.01). CONCLUSION: The present results suggested that protective effect of SME against CCl4-induced hepatic fibrosis may rely on its effect on reducing oxidative stress and improving drug metabolizing enzyme activity in liver.

Expression of C-peptide multiple gene copies in Escherichia coli and stabilities of C-peptide in aqueous solution.

A gene fragment encoding three copies of proinsulin C-peptide was synthesized and expressed in E. coli and the recombinant proinsulin C-peptide was produced through site-specific cleavage of the resulting gene products. The fusion protein was expressed at high level, about 80 mg/L, as a soluble product in the cytoplasm. Ni-NTA affinity chromatography efficiently separated the expressed fusion protein from the supernatant, to obtain about 37.5 mg/L of the fusion protein with 70% purity. Enzymatic digestion by trypsin and carboxypeptidase B of the fusion protein efficiently released native C-peptide, the overall yield of recombinant C-peptide at a purity over 95% was 1.5 mg/L. The good agreement of amino acids composition, together with shown similarities of the recombinant C-peptide to C-peptide standard in the comparative RP-HPLC analysis and IMMULITE C-Peptide quantitative assay, suggested that the recombinant C-peptide obtained in this report was the native human C-peptide. The investigation of the chemical stability of recombinant human C-peptide in aqueous solutions by RP-HPLC was also reported. The degradation of the recombinant C-peptide showed a marked dependence on pH and temperature. The degradation reaction of C-peptide occurred immediately in pH 3 or pH 9 buffered solution. The degradation reaction of C-peptide followed first-order kinetics in pH 3 buffered solution at 37 degrees C or 70 degrees C, only 40.3% of C-peptide was remained after 10 h at 70 degrees C. The maximum stability was achieved at pH 7.4, more than 90% of C-peptide were detected at pH 7.4 and 37 degrees C after 10 h and at pH 7.4 and 70 degrees C after 5 h. 99% and 96% of C-peptide was remained at pH 7.4 and 37 degrees C after 10 h with and without 10 g/L BSA respectively.

State observer design for delayed genetic regulatory networks.

Genetic regulatory networks are dynamic systems which describe the interactions among gene products (mRNAs and proteins). The internal states of a genetic regulatory network consist of the concentrations of mRNA and proteins involved in it, which are very helpful in understanding its dynamic behaviors. However, because of some limitations such as experiment techniques, not all internal states of genetic regulatory network can be effectively measured. Therefore it becomes an important issue to estimate the unmeasured states via the available measurements. In this study, we design a state observer to estimate the states of genetic regulatory networks with time delays from available measurements. Furthermore, based on linear matrix inequality (LMI) approach, a criterion is established to guarantee that the dynamic of estimation error is globally asymptotically stable. A gene repressillatory network is employed to illustrate the effectiveness of our design approach.

[Prospective study on comparing two methods: for dynamic monitoring of minimal residual disease in acute promyelocytic leukemia].

This study was purposed to comparatively analyse the value of PCR and FCM for dynamic monitoring minimal residual disease (MRD) of acute promyelocytic leukemia. The patients with acute promyelocytic leukemia hospitalized in our hospital from January 2011 to December 2012 were observed and all achieved complete remission after remission induction therapy. Before the chemotherapy, the bone marrow cell morphology examination, polymerase-chain reaction (PCR) and multi-parameter flow cytometry (FCM) were performed for each patient. Then the detection results were statistically analyzed. The 477 specimens were achieved from 159 detections for 48 patients. The results showed that 3 specimens were found to be relapsed by bone marrow cell examination, and other specimens were complete remission;PCR detection confirmed 7 positive, and the FCM confirmed 19 positive. There wasn't significant difference between PCR and FCM by kappa test (P > 0.05). It is concluded that FCM is as sensitive as PCR in evaluating the treatment effect of acute promyelocytic leukemia.

Influence of adriamycin on changes in Nanog, Oct-4, Sox2, ARID1 and Wnt5b expression in liver cancer stem cells.

AIM: To determine the influence of Adriamycin (ADM) on the changes in Nanog, Oct4, Sox2, as well as, in ARID1 and Wnt5b expression in liver cancer stem cells. METHODS: The MHCC97-L and HCCLM3 liver cancer cell lines were selected as the cell models in this study, and were routinely cultured. The 50% lethal dose (LD50) in the cell lines was detected by the MTT assay. Expression changes in liver cancer stem cell related genes (Nanog, Oct-4, Sox2, ARID1, and Wnt5b) were detected by western blot following treatment with ADM (LD50). RESULTS: The LD50 of ADM in MHCC97-L cells was lower than that in HCCLM3 cells (0.4123 ± 0.0236 µmol/L vs 0.5259 ± 0.0125 µmol/L, P < 0.05). Wnt5b and Nanog were expressed in both MHCC97-L and HCCLM3 cells, while only Sox2 was expressed in HCCLM3 cells. However, neither ARID1A nor Oct4 was detected in these two cell lines. Genes, related to the stem cells, showed different expression in liver cancer cells with different metastatic potential following treatment with ADM (LD50). Wnt5b protein increased gradually within 4 h of ADM (LD50) treatment, while Nanog decreased (P < 0.05). After 12 h, Wnt5b decreased gradually, while Nanog increased steadily (P < 0.05). In addition, only Sox2 was expressed in HCCLM3 cells with high metastatic potential following ADM (LD50) treatment. The expression of Sox2 increased gradually with ADM (LD50) in HCCLM3 cells (P < 0.05). CONCLUSION: ADM increased the death rate of MHCC97-L and HCCLM3 cells, while the growth suppressive effect of ADM was higher in MHCC97-L cells than in HCCLM3 cells.

Complexity analysis and parameter estimation of dynamic metabolic systems.

A metabolic system consists of a number of reactions transforming molecules of one kind into another to provide the energy that living cells need. Based on the biochemical reaction principles, dynamic metabolic systems can be modeled by a group of coupled differential equations which consists of parameters, states (concentration of molecules involved), and reaction rates. Reaction rates are typically either polynomials or rational functions in states and constant parameters. As a result, dynamic metabolic systems are a group of differential equations nonlinear and coupled in both parameters and states. Therefore, it is challenging to estimate parameters in complex dynamic metabolic systems. In this paper, we propose a method to analyze the complexity of dynamic metabolic systems for parameter estimation. As a result, the estimation of parameters in dynamic metabolic systems is reduced to the estimation of parameters in a group of decoupled rational functions plus polynomials (which we call improper rational functions) or in polynomials. Furthermore, by taking its special structure of improper rational functions, we develop an efficient algorithm to estimate parameters in improper rational functions. The proposed method is applied to the estimation of parameters in a dynamic metabolic system. The simulation results show the superior performance of the proposed method.

M-matrix-based stability conditions for genetic regulatory networks with time-varying delays and noise perturbations.

Stability is essential for designing and controlling any dynamic systems. Recently, the stability of genetic regulatory networks has been widely studied by employing linear matrix inequality (LMI) approach, which results in checking the existence of feasible solutions to high-dimensional LMIs. In the previous study, the authors present several stability conditions for genetic regulatory networks with time-varying delays, based on M-matrix theory and using the non-smooth Lyapunov function, which results in determining whether a low-dimensional matrix is a non-singular M-matrix. However, the previous approach cannot be applied to analyse the stability of genetic regulatory networks with noise perturbations. Here, the authors design a smooth Lyapunov function quadratic in state variables and employ M-matrix theory to derive new stability conditions for genetic regulatory networks with time-varying delays. Theoretically, these conditions are less conservative than existing ones in some genetic regulatory networks. Then the results are extended to genetic regulatory networks with time-varying delays and noise perturbations. For genetic regulatory networks with n genes and n proteins, the derived conditions are to check if an n × n matrix is a non-singular M-matrix. To further present the new theories proposed in this study, three example regulatory networks are analysed.

Robust and global delay-dependent stability for genetic regulatory networks with parameter uncertainties.

The study of stability is essential for designing or controlling genetic regulatory networks, which can be described by nonlinear differential equations with time delays. Much attention has been paid to the study of delay-independent stability of genetic regulatory networks and as a result, many sufficient conditions have been derived for delay-independent stability. Although it might be more interesting in practice, delay-dependent stability of genetic regulatory networks has been studied insufficiently. Based on the linear matrix inequality (LMI) approach, in this study we will present some delay-dependent stability conditions for genetic regulatory networks. Then we extend these results to genetic regulatory networks with parameter uncertainties. To illustrate the effectiveness of our theoretical results, gene repressilatory networks are analyzed .