A Pseudo-EM Algorithm for Clustering Incomplete Longitudinal Data

Thyroid hormone inhibits slow skeletal TnI expression in cardiac TnI-null myocardial cells

BackgroundThe vascular disease in-stent restenosis (ISR) is characterized by formation of neointima and adverse inward remodeling of the artery after injury by coronary stent implantation. We hypothesized that the analysis of gene expression in peripheral blood mononuclear cells (PBMCs) would demonstrate differences in transcript expression between individuals who develop ISR and those who do not.Methods and ResultsWe determined and investigated PBMC gene expression of 358 patients undergoing an index procedure to treat in de novo coronary artery lesions with bare metallic stents, using a novel time-varying intercept model to optimally assess the time course of gene expression across a time course of blood samples. Validation analyses were conducted in an independent sample of 97 patients with similar time-course blood sampling and gene expression data. We identified 47 probesets with differential expression, of which 36 were validated upon independent replication testing. The genes identified have varied functions, including some related to cellular growth and metabolism, such as the NAB2 and LAMP genes.ConclusionsIn a study of patients undergoing bare metallic stent implantation, we have identified and replicated differential gene expression in peripheral blood mononuclear cells, studied across a time series of blood samples. The genes identified suggest alterations in cellular growth and metabolism pathways, and these results provide the basis for further specific functional hypothesis generation and testing of the mechanisms of ISR.

Studying the time course of gene expression in injured skeletal muscle would help to estimate the timing of injuries. In this study, we investigated large-scale gene expression in incision-injured mouse skeletal muscle by DNA microarray using correspondence analysis (CA). Biceps femoris muscle samples were collected 6, 12, and 24 hours after injury, and RNA was extracted and prepared for microarray analysis. On a 2-dimensional plot by CA, the genes (row score coordinate) located farther from each time series (column score coordinate) had more upregulation at particular times. Each gene was situated in 6 subdivided triangular areas according to the magnitude of the relationship of the fold change (FC) value at each time point compared to the control. In each area, genes for which the ratios of two particular FC values were close to 1 were distributed along the two border lines. There was a tendency for genes whose FC values were almost equal to be distributed near the intersection of these 6 areas. Therefore, the gene marker candidates for estimation of the timing of injuries were detectable according to the location on the CA plot. Moreover, gene sets created by a specific gene and its surrounding genes were composed of genes that showed similar or identical fluctuation patterns to the specific gene. In various analyses on these sets, significant gene ontology term and pathway activity may reflect changes in specific genes. In conclusion, analyses of gene sets based on CA plots is effective for investigation of the time-dependent fluctuation in gene expression after injury.

Gene expression time course data can be used not only to detect differentially expressed genes but also to find temporal associations among genes. The problem of reconstructing generalized logical networks to account for temporal dependencies among genes and environmental stimuli from transcriptomic data is addressed. A network reconstruction algorithm was developed that uses statistical significance as a criterion for network selection to avoid false-positive interactions arising from pure chance. The multinomial hypothesis testing-based network reconstruction allows for explicit specification of the false-positive rate, unique from all extant network inference algorithms. The method is superior to dynamic Bayesian network modeling in a simulation study. Temporal gene expression data from the brains of alcohol-treated mice in an analysis of the molecular response to alcohol are used for modeling. Genes from major neuronal pathways are identified as putative components of the alcohol response mechanism. Nine of these genes have associations with alcohol reported in literature. Several other potentially relevant genes, compatible with independent results from literature mining, may play a role in the response to alcohol. Additional, previously unknown gene interactions were discovered that, subject to biological verification, may offer new clues in the search for the elusive molecular mechanisms of alcoholism.

BackgroundThe analysis of gene expression from time series underpins many biological studies. Two basic forms of analysis recur for data of this type: removing inactive (quiet) genes from the study and determining which genes are differentially expressed. Often these analysis stages are applied disregarding the fact that the data is drawn from a time series. In this paper we propose a simple model for accounting for the underlying temporal nature of the data based on a Gaussian process.ResultsWe review Gaussian process (GP) regression for estimating the continuous trajectories underlying in gene expression time-series. We present a simple approach which can be used to filter quiet genes, or for the case of time series in the form of expression ratios, quantify differential expression. We assess via ROC curves the rankings produced by our regression framework and compare them to a recently proposed hierarchical Bayesian model for the analysis of gene expression time-series (BATS). We compare on both simulated and experimental data showing that the proposed approach considerably outperforms the current state of the art.ConclusionsGaussian processes offer an attractive trade-off between efficiency and usability for the analysis of microarray time series. The Gaussian process framework offers a natural way of handling biological replicates and missing values and provides confidence intervals along the estimated curves of gene expression. Therefore, we believe Gaussian processes should be a standard tool in the analysis of gene expression time series.

Lipid-mediated delivery of RNA is more efficient than delivery of DNA in non-dividing cells

Occlusion of two adjacent cortical veins is followed by a widespread reduction of rCBF and the occurrence of small infarcts, which become larger if spreading depression (SD) occurs. The infarct matures over time with TUNEL-positive cells seen in the penumbra up to 4 days after vein occlusion. Caspase inhibition with zVAD.fmk reduces infarct size. Here, the time course of gene expression in the penumbra is compared to that induced by SD alone.

The particle bombardment (gene gun) method for gene transfer provides a new and efficient means for transfection of various cell types in culture. In this study we evaluate its application to human prostate tumor cells. Transient expression of the firefly luciferase gene driven by five viral and five cellular promoters was assessed after in vitro gene transfer using the gene gun method. The relative strengths of these promoters were quantitatively determined in four different human prostate tumor cell lines: DU145, PC-3, LNCaP, and CWR22Rv1 cells. In situ histochemical staining of cells, transfected with bacterial beta-galactosidase cDNA as a reporter gene, was also performed to evaluate the transfection efficiency. Time course of gene expression was determined using the luciferase reporter gene. The peak levels of transient expression of firefly luciferase are observed within 24 hr after gene transfer. Sustained but reduced luciferase levels were also detected as long as 5 days post transfection. Up to 35% of bombarded cells in vitro were found to express transgenic beta-galactosidase activity. Among tested viral promoters, cytomegalovirus early enhancer/promoter activity was observed to confer consistently the highest activity in each test cell line, whereas phosphoglycerate kinase gene promoter possessed the highest activity among the cellular promoters tested. The particle bombardment gene-transfer technology can be effectively employed as an efficient method for in vitro gene-transfer into prostate tumor cells. The characterization of relative promoter strength and preference may be useful for future studies of cancer gene therapy approaches.

Gene therapy for heart diseases requires availability of an efficient vector for gene transfer into myocardium. Recombinant adenovirus expressing the Escherichia coli beta-galactosidase (beta-Gal) gene was shown to infect rat cardiocytes efficiently in vivo. However, a time course of gene expression showed that transgene expression was maximal during the first week following injection, then declined and disappeared by day 21. An immunosuppressive treatment prolonged beta-Gal expression for at least 21 days. On the contrary, a preimmunization of the animals by two intraperitoneal injections of the vector led to a decreased transgene expression 48 hr after intramyocardial injection and to a barely detectable expression at the sixth day. Appearance of adenovirus neutralizing antibodies in preimmunized animals could have contributed to such a refractoriness to further adenoviral infection. Finally, a neonatal intrathymic injection of the vector was able to induce long-term LacZ expression for more than 2 months after heart injection, although neutralizing as well as anti-beta-Gal antibodies were detected in sera of the animals. These results indicate that an immune response against first-generation replication-defective adenoviral vectors is a major cause of transient transgene expression, a cellular response being most probably responsible for ablation of transgene expression in immunocompetent animals.

To clarify the acquisition of the adult muscle pattern in Xenopus laevis, in situ hybridization and reverse transcriptase-polymerase chain reaction were used to correlate the time course of gene expression for myogenic regulatory factors (Myf-5, MyoD, and myogenin) with the expression of contractile protein (myosin heavy chain; MHC) genes during hindlimb formation compared with their expression in dorsal body muscles. After the precocious expression of Myf-5 and MyoD mRNA in limb bud (stage 50), myogenin mRNA strongly accumulated later at paddle stages (stages 52/53) concomitantly with the accumulation of both the larval and the adult MHC mRNAs. In dorsal body muscles, as early as stage 52, myogenin transcripts accumulated in a few small, secondary myofibers expressing the adult MHC mRNA that were located along the dorsomedial edge, but they were never detected in the large, primary myofibers of the body expressing the larval MHC mRNA. During metamorphosis, the areas expressing both the adult MHC and the myogenin transcripts gradually expanded from the dorsomedial edge to the ventral side of the dorsal body muscles, accounting for the progression of the secondary "adult" myogenesis described previously (Nishikawa and Hayashi [1994] Dev. Biol. 165:86-94). This work shows that, in Xenopus, the accumulation of myogenin mRNA is restricted to secondary myogenesis, including the formation of new muscles in developing limbs as well as in dorsal muscles during body remodeling. This shows that myogenin is not required for primary myogenesis, and it suggests a crucial role for myogenin in the terminal differentiation program, including myoblast fusion and the activation of adult-type muscle genes.

As a first step to creating a comprehensive atlas of gene-expression patterns during Drosophila embryogenesis, 2,179 genes have been examinded by in situ hybridization to fixed Drosophila embryos. Of the genes assayed, 63.7% displayed dynamic expression patterns that were documented with 25,690 digital photomicrographs of individual embryos.

Time course, localization and pharmacological modulation of immediate early inducible genes, brain-derived neurotrophic factor and trkB messenger RNAs in the rat brain following photochemical stroke

Time course of hepatic gene expression and plasma vitellogenin protein concentrations in estrone-exposed juvenile rainbow trout (Oncorhynchus mykiss)

The mycotoxin ochratoxin A (OTA) is a potent nephrocarcinogen, mainly in male rats. The aim of this study was to determine the time course of gene expression (GeneChip® Rat Gene 2.0 ST Array, Affymetrix) in kidney samples from male and female F344 rats, treated daily (p.o) with 0.50 mg/kg b.w. (body weight) of OTA for 7 or 21 days, and evaluate if there were differences between both sexes. After OTA treatment, there was an evolution of gene expression in the kidney over time, with more differentially expressed genes (DEG) at 21 days. The gene expression time course was different between sexes with respect to the number of DEG and the direction of expression (up or down): the female response was progressive and consistent over time, whereas males had a different early response with more DEG, most of them up-regulated. The statistically most significant DEG corresponded to metabolism enzymes (Akr1b7, Akr1c2, Adh6 down-regulated in females; Cyp2c11, Dhrs7, Cyp2d1, Cyp2d5 down-regulated in males) or transporters (Slc17a9 down-regulated in females; Slco1a1 (OATP-1) and Slc51b and Slc22a22 (OAT) down-regulated in males). Some of these genes had also a basal sex difference and were over-expressed in males or females with respect to the other sex.

Data clustering is one of the popular tasks recently used in the educational data mining arena for grouping similar students by several aspects such as study performance, behavior, skill, etc. Many well-known clustering algorithms such as k-means, expectation-maximization, spectral clustering, etc. were employed in the related works. None of them has taken into consideration the incompleteness of the educational data gathered in an academic credit system. If just a few records have missing values, we might ignore them in the mining task. However, as there are a large number of missing values, ignoring them may lead to the data insufficiency and ineffectiveness of the mining task. Hence, we define a robust and effective algorithmic framework for incomplete educational data clustering using the nearest prototype strategy. Within the framework, we propose two novel incomplete educational data clustering algorithms K_nps and S_nps based on the k-means algorithm and the self-organizing map, respectively. Experimental results have shown that the clusters from our proposed algorithms have better cluster quality as compared to the different existing approaches.