Ann M. Hess

Linear trend analysis: a comparison of methods

In this paper, we present an overview of statistical approaches available for detecting and estimating linear trends in environmental data. We evaluate seven methods of trend detection and make recommendations based on a simulation study. We also illustrate the methods using real data.

Filtering for increased power for microarray data analysis

BackgroundDue to the large number of hypothesis tests performed during the process of routine analysis of microarray data, a multiple testing adjustment is certainly warranted. However, when the number of tests is very large and the proportion of differentially expressed genes is relatively low, the use of a multiple testing adjustment can result in very low power to detect those genes which are truly differentially expressed. Filtering allows for a reduction in the number of tests and a corresponding increase in power. Common filtering methods include filtering by variance, average signal or MAS detection call (for Affymetrix arrays). We study the effects of filtering in combination with the Benjamini-Hochberg method for false discovery rate control and q-value for false discovery rate estimation.ResultsThree case studies are used to compare three different filtering methods in combination with the two false discovery rate methods and three different preprocessing methods. For the case studies considered, filtering by detection call and variance (on the original scale) consistently led to an increase in the number of differentially expressed genes identified. On the other hand, filtering by variance on the log2 scale had a detrimental effect when paired with MAS5 or PLIER preprocessing methods, even when the testing was done on the log2 scale. A simulation study was done to further examine the effect of filtering by variance. We find that filtering by variance leads to higher power, often with a decrease in false discovery rate, when paired with either of the false discovery rate methods considered. This holds regardless of the proportion of genes which are differentially expressed or whether we assume dependence or independence among genes.ConclusionThe case studies show that both detection call and variance filtering are viable methods of filtering which can increase the number of differentially expressed genes identified. The simulation study demonstrates that when paired with a false discovery rate method, filtering by variance can increase power while still controlling the false discovery rate. Filtering out 50% of probe sets seems reasonable as long as the majority of genes are not expected to be differentially expressed.

Comparative genomics of small RNA regulatory pathway components in vector mosquitoes

BackgroundSmall RNA regulatory pathways (SRRPs) control key aspects of development and anti-viral defense in metazoans. Members of the Argonaute family of catalytic enzymes degrade target RNAs in each of these pathways. SRRPs include the microRNA, small interfering RNA (siRNA) and PIWI-type gene silencing pathways. Mosquitoes generate viral siRNAs when infected with RNA arboviruses. However, in some mosquitoes, arboviruses survive antiviral RNA interference (RNAi) and are transmitted via mosquito bite to a subsequent host. Increased knowledge of these pathways and functional components should increase understanding of the limitations of anti-viral defense in vector mosquitoes. To do this, we compared the genomic structure of SRRP components across three mosquito species and three major small RNA pathways.ResultsThe Ae. aegypti, An. gambiae and Cx. pipiens genomes encode putative orthologs for all major components of the miRNA, siRNA, and piRNA pathways. Ae. aegypti and Cx. pipiens have undergone expansion of Argonaute and PIWI subfamily genes. Phylogenetic analyses were performed for these protein families. In addition, sequence pattern recognition algorithms MEME, MDScan and Weeder were used to identify upstream regulatory motifs for all SRRP components. Statistical analyses confirmed enrichment of species-specific and pathway-specific cis-elements over the rest of the genome.ConclusionAnalysis of Argonaute and PIWI subfamily genes suggests that the small regulatory RNA pathways of the major arbovirus vectors, Ae. aegypti and Cx. pipiens, are evolving faster than those of the malaria vector An. gambiae and D. melanogaster. Further, protein and genomic features suggest functional differences between subclasses of PIWI proteins and provide a basis for future analyses. Common UCR elements among SRRP components indicate that 1) key components from the miRNA, siRNA, and piRNA pathways contain NF-kappaB-related and Broad complex transcription factor binding sites, 2) purifying selection has occurred to maintain common pathway-specific elements across mosquito species and 3) species-specific differences in upstream elements suggest that there may be differences in regulatory control among mosquito species. Implications for arbovirus vector competence in mosquitoes are discussed.

Small RNA profiling of Dengue virus-mosquito interactions implicates the PIWI RNA pathway in anti-viral defense

BackgroundSmall RNA (sRNA) regulatory pathways (SRRPs) are important to anti-viral defence in mosquitoes. To identify critical features of the virus infection process in Dengue serotype 2 (DENV2)-infected Ae. aegypti, we deep-sequenced small non-coding RNAs. Triplicate biological replicates were used so that rigorous statistical metrics could be applied.ResultsIn addition to virus-derived siRNAs (20-23 nts) previously reported for other arbovirus-infected mosquitoes, we show that PIWI pathway sRNAs (piRNAs) (24-30 nts) and unusually small RNAs (usRNAs) (13-19 nts) are produced in DENV-infected mosquitoes. We demonstrate that a major catalytic enzyme of the siRNA pathway, Argonaute 2 (Ago2), co-migrates with a ~1 megadalton complex in adults prior to bloodfeeding. sRNAs were cloned and sequenced from Ago2 immunoprecipitations. Viral sRNA patterns change over the course of infection. Host sRNAs were mapped to the published aedine transcriptome and subjected to analysis using edgeR (Bioconductor). We found that sRNA profiles are altered early in DENV2 infection, and mRNA targets from mitochondrial, transcription/translation, and transport functional categories are affected. Moreover, small non-coding RNAs (ncRNAs), such as tRNAs, spliceosomal U RNAs, and snoRNAs are highly enriched in DENV-infected samples at 2 and 4 dpi.ConclusionsThese data implicate the PIWI pathway in anti-viral defense. Changes to host sRNA profiles indicate that specific cellular processes are affected during DENV infection, such as mitochondrial function and ncRNA levels. Together, these data provide important progress in understanding the DENV2 infection process in Ae. aegypti.

Transcriptome analyses give insights into selenium-stress responses and selenium tolerance mechanisms in Arabidopsis.

Selenate is chemically similar to sulfate and can be taken up and assimilated by plants via the same transporters and enzymes. In contrast to many other organisms, selenium (Se) has not been shown to be essential for higher plants. In excess, Se is toxic and restricts development. Both Se deficiency and toxicity pose problems worldwide. To obtain better insights into the effects of Se on plant metabolism and into plant mechanisms involved in Se tolerance, the transcriptome of Arabidopsis plants grown with or without selenate was studied and Se-responsive genes identified. Roots and shoots exhibited different Se-related changes in gene regulation and metabolism. Many genes involved in sulfur (S) uptake and assimilation were upregulated. Accordingly, Se treatment enhanced sulfate levels in plants, but the quantity of organic S metabolites decreased. Transcripts regulating the synthesis and signaling of ethylene and jasmonic acid were also upregulated by Se. Arabidopsis mutants defective in ethylene or jasmonate response pathways exhibited reduced tolerance to Se, suggesting an important role for these two stress hormones in Se tolerance. Selenate upregulated a variety of transcripts that were also reportedly induced by salt and osmotic stress. Selenate appeared to repress plant development, as suggested by the downregulation of genes involved in cell wall synthesis and auxin-regulated proteins. The Se-responsive genes discovered in this study may help create plants that can better tolerate and accumulate Se, which may enhance the effectiveness of Se phytoremediation or serve as Se-fortified food.

Fisher's combined p-value for detecting differentially expressed genes using Affymetrix expression arrays

BackgroundCurrently, most tests of differential gene expression using Affymetrix expression array data are performed using expression summary values representing each probe set on a microarray. Recently testing methods have been proposed which incorporate probe level information. We propose a new approach that uses Fishers method of combining evidence from multiple sources of information. Specifically, we combine p-values from probe level tests of significance.ResultsThe combined p method and other competing methods were compared using three spike-in datasets (where probe sets corresponding differentially spiked transcripts are known) and array data from a biological study validated with qRT-PCR. Based on power and false discovery rates computed for the spike-in datasets, we demonstrate that the combined p method compares favorably with other methods. We find that probe level testing methods select many of the same genes as differentially expressed. We illustrate the use of the combined p method for diagnostic purposes using examples.ConclusionCombined p is a promising alternative to existing methods of testing for differential gene expression. In addition, the combined p method is particularly well suited as a diagnostic tool for exploratory analysis of microarray data.

MicroRNA levels are modulated in Aedes aegypti after exposure to Dengue-2

To define microRNA (miRNA) involvement during arbovirus infection of Aedes aegypti, we mined deep sequencing libraries of Dengue type 2 (DENV2)‐exposed mosquitoes. Three biological replicates for each timepoint [2, 4 and 9 days post‐exposure (dpe)] and treatment group allowed us to remove the outliers associated with sample‐to‐sample variability. Using edgeR (R Bioconductor), designed for use with replicate deep sequencing data, we determined the log fold‐change (logFC) of miRNA levels [18–23 nucleotides (nt)]. The number of significantly modulated miRNAs increased from ≤5 at 2 and 4 dpe to 23 unique miRNAs by 9 dpe. Putative miRNA targets were predicted by aligning miRNAs to the transcriptome, and the list was reduced to include the intersection of hits found using the Miranda, PITA, and TargetScan algorithms. To further reduce false‐positives, putative targets were validated by cross‐checking them with mRNAs reported in recent DENV2 host response transcriptome reports; 4076 targets were identified. Of these, 464 gene targets have predicted miRNA‐binding sites in 3′ untranslated regions. Context‐specific target functional groups include proteins involved in transport, transcriptional regulation, mitochondrial function, chromatin modification and signal transduction processes known to be required for viral replication and dissemination. The miRNA response is placed in context with other vector host response studies by comparing the predicted targets with those of transcriptome studies. Together, these data are consistent with the hypothesis that profound and persistent changes to gene expression occur in DENV2‐exposed mosquitoes.

A metabolic biosignature of early response to anti-tuberculosis treatment

BackgroundThe successful treatment of tuberculosis (TB) requires long-term multidrug chemotherapy. Clinical trials to evaluate new drugs and regimens for TB treatment are protracted due to the slow clearance of Mycobacterium tuberculosis (Mtb) infection and the lack of early biomarkers to predict treatment outcome. Advancements in the field of metabolomics make it possible to identify metabolic profiles that correlate with disease states or successful chemotherapy. However, proof-of-concept of this approach has not been provided for a TB-early treatment response biosignature (TB-ETRB).MethodsUrine samples collected at baseline and during treatment from 48 Ugandan and 39 South African HIV-seronegative adults with pulmonary TB were divided into discovery and qualification sets, normalized to creatinine concentration, and analyzed by liquid chromatography-mass spectrometry to identify small molecule molecular features (MFs) in individual patient samples. A biosignature that distinguished baseline and 1 month treatment samples was selected by pairwise t-test using data from two discovery sample sets. Hierarchical clustering and repeated measures analysis were applied to additional sample data to down select molecular features that behaved consistently between the two clinical sites and these were evaluated by logistic regression analysis.ResultsAnalysis of discovery samples identified 45 MFs that significantly changed in abundance at one month of treatment. Down selection using an extended set of discovery samples and qualification samples confirmed 23 MFs that consistently changed in abundance between baseline and 1, 2 and 6 months of therapy, with 12 MFs achieving statistical significance (p < 0.05). Six MFs classified the baseline and 1 month samples with an error rate of 11.8%.ConclusionsThese results define a urine based TB-early treatment response biosignature (TB-ETRB) applicable to different parts of Africa, and provide proof-of-concept for further evaluation of this technology in monitoring clinical responses to TB therapy.

Descriptive proteomic analysis shows protein variability between closely related clinical isolates of Mycobacterium tuberculosis

The use of isobaric tags such as iTRAQ allows the relative and absolute quantification of hundreds of proteins in a single experiment for up to eight different samples. More classical techniques such as 2‐DE can offer a complimentary approach for the analysis of complex protein samples. In this study, the proteomes of secreted and cytosolic proteins of genetically closely related strains of Mycobacterium tuberculosis were analyzed. Analysis of 2‐D gels afforded 28 spots with variations in protein abundance between strains. These were identified by MS/MS. Meanwhile, a rigorous statistical analysis of iTRAQ data allowed the identification and quantification of 101 and 137 proteins in the secreted and cytosolic fractions, respectively. Interestingly, several differences in protein levels were observed between the closely related strains BE, C28 and H6. Seven proteins related to cell wall and cell processes were more abundant in BE, while enzymes related to metabolic pathways (GltA2, SucC, Gnd1, Eno) presented lower levels in the BE strain. Proteins involved in iron and sulfur acquisition (BfrB, ViuB, TB15.3 and SseC2) were more abundant in C28 and H6. In general, iTRAQ afforded rapid identification of fine differences between protein levels such as those presented between closely related strains. This provides a platform from which the relevance of these differences can be assessed further using complimentary proteomic and biological modeling methods.

Detection of genomic deletions in rice using oligonucleotide microarrays

BackgroundThe induction of genomic deletions by physical- or chemical- agents is an easy and inexpensive means to generate a genome-saturating collection of mutations. Different mutagens can be selected to ensure a mutant collection with a range of deletion sizes. This would allow identification of mutations in single genes or, alternatively, a deleted group of genes that might collectively govern a trait (e.g., quantitative trait loci, QTL). However, deletion mutants have not been widely used in functional genomics, because the mutated genes are not tagged and therefore, difficult to identify. Here, we present a microarray-based approach to identify deleted genomic regions in rice mutants selected from a large collection generated by gamma ray or fast neutron treatment. Our study focuses not only on the utility of this method for forward genetics, but also its potential as a reverse genetics tool through accumulation of hybridization data for a collection of deletion mutants harboring multiple genetic lesions.ResultsWe demonstrate that hybridization of labeled genomic DNA directly onto the Affymetrix Rice GeneChip® allows rapid localization of deleted regions in rice mutants. Deletions ranged in size from one gene model to ~500 kb and were predicted on all 12 rice chromosomes. The utility of the technique as a tool in forward genetics was demonstrated in combination with an allelic series of mutants to rapidly narrow the genomic region, and eventually identify a candidate gene responsible for a lesion mimic phenotype. Finally, the positions of mutations in 14 mutants were aligned onto the rice pseudomolecules in a user-friendly genome browser to allow for rapid identification of untagged mutations http://irfgc.irri.org/cgi-bin/gbrowse/IR64_deletion_mutants/.ConclusionWe demonstrate the utility of oligonucleotide arrays to discover deleted genes in rice. The density and distribution of deletions suggests the feasibility of a database saturated with deletions across the rice genome. This community resource can continue to grow with further hybridizations, allowing researchers to quickly identify mutants that harbor deletions in candidate genomic regions, for example, regions containing QTL of interest.