Post hoc pattern matching: assigning significance to statistically defined expression patterns in single channel microarray data

HPLC-MS Analysis of Lichen-Derived Metabolites in the Life Stages of Crambidia cephalica (Grote & Robinson)

Tiger moths (Lepidoptera: Erebidae: Arctiinae: Arctiini) are notable for their specialized associations with hosts that produce toxic secondary compounds, and are thus an ideal study system for understanding insect-plant interactions and the evolution of antipredatory defense. Likewise, their sister lineage (Arctiinae: Lithosiini) has been documented feeding on algae and lichens, and is known to sequester lichen-derived secondary compounds from the larval to adult stages. Prevalence of lichenivory in this early radiation (ca. 3000 species) may provide clues to the phylogenetic basis for storied chemical sequestration within all tiger moths. Despite the evolutionary significance of this trait, we lack a basic understanding of the extent of lichenivory among lithosiines, and the distribution of sequestered chemicals among life stages. The dynamics of chemical sequestration throughout the lifecycle for the lichen moth Crambidia cephalica were investigated by testing the hypothesis that lichen-derived metabolites are unequally distributed among life stages, and that laboratory-reared C. cephalica have less metabolite diversity than wild-caught individuals. Crambidia cephalica was reared on Physcia, and analyzed using high-performance liquid chromatography coupled to mass spectrometry (HPLC-MS). Several putative lichen-derived metabolites were detected across three life stages, i.e., larval, pupal, and adult, and differences among life stages and lichen host were observed. These results provide evidence that multiple lichen-derived metabolites are sequestered by C. cephalica; some metabolites are retained through adulthood, and others are lost or modified in earlier life stages. The presence of differing lichen-derived metabolites across life stages may indicate functional properties of the metabolites for C. cephalica with regards to chemical protection from antagonists, and other physiological processes.

Thermodynamically optimal whole-genome tiling microarray design and validation

Background

Microarray is an efficient apparatus to interrogate the whole transcriptome of species. Microarray can be designed according to annotated gene sets, but the resulted microarrays cannot be used to identify novel transcripts and this design method is not applicable to unannotated species. Alternatively, a whole-genome tiling microarray can be designed using only genomic sequences without gene annotations, and it can be used to detect novel RNA transcripts as well as known genes. The difficulty with tiling microarray design lies in the tradeoff between probe-specificity and coverage of the genome. Sequence comparison methods based on BLAST or similar software are commonly employed in microarray design, but they cannot precisely determine the subtle thermodynamic competition between probe targets and partially matched probe nontargets during hybridizations.

Findings

Using the whole-genome thermodynamic analysis software PICKY to design tiling microarrays, we can achieve maximum whole-genome coverage allowable under the thermodynamic constraints of each target genome. The resulted tiling microarrays are thermodynamically optimal in the sense that all selected probes share the same melting temperature separation range between their targets and closest nontargets, and no additional probes can be added without violating the specificity of the microarray to the target genome.

Conclusions

This new design method was used to create two whole-genome tiling microarrays for Escherichia coli MG1655 and Agrobacterium tumefaciens C58 and the experiment results validated the design.

Electronic supplementary material

The online version of this article (doi:10.1186/s13104-016-2113-4) contains supplementary material, which is available to authorized users.

Exploiting dependencies of pairwise comparison outcomes to predict patterns of gene response

BACKGROUND

The analysis of gene expression has played an important role in medical and bioinformatics research. Although it is known that a large number of samples is needed to determine the patterns of gene expression accurately, practical designs of gene expression studies occasionally have insufficient numbers of samples, making it difficult to ascertain true response patterns of variantly expressed genes.

RESULTS

We describe an approach to cope with the challenge of predicting true orders of gene response to treatments. We show that true patterns of gene response must be orderable sets. In experiments with few samples, we modify the conventional pairwise comparison tests and increase the significance level α intelligently to deduce orderable patterns, which are most likely true orders of gene response. Additionally, motivated by the fact that a gene can be involved in multiple biological functions, our method further resamples experimental replicates and predicts multiple response patterns for each gene.

CONCLUSIONS

This method can be useful in designing cost-effective experiments with small sample sizes. Patterns of highly-variantly expressed genes can be predicted by varying α intelligently. Furthermore, clusters are labeled meaningfully with patterns that describe precisely how genes in such clusters respond to treatments.

Discovery of Emphysema Relevant Molecular Networks from an A/J Mouse Inhalation Study Using Reverse Engineering and Forward Simulation (REFS™)

Chronic obstructive pulmonary disease (COPD) is a respiratory disorder caused by extended exposure of the airways to noxious stimuli, principally cigarette smoke (CS). The mechanisms through which COPD develops are not fully understood, though it is believed that the disease process includes a genetic component, as not all smokers develop COPD. To investigate the mechanisms that lead to the development of COPD/emphysema, we measured whole genome gene expression and several COPD-relevant biological endpoints in mouse lung tissue after exposure to two CS doses for various lengths of time. A novel and powerful method, Reverse Engineering and Forward Simulation (REFS™), was employed to identify key molecular drivers by integrating the gene expression data and four measured COPD-relevant endpoints (matrix metalloproteinase (MMP) activity, MMP-9 levels, tissue inhibitor of metalloproteinase-1 levels and lung weight). An ensemble of molecular networks was generated using REFS™, and simulations showed that it could successfully recover the measured experimental data for gene expression and COPD-relevant endpoints. The ensemble of networks was then employed to simulate thousands of in silico gene knockdown experiments. Thirty-three molecular key drivers for the above four COPD-relevant endpoints were therefore identified, with the majority shown to be enriched in inflammation and COPD.

External light conditions and internal cell cycle phases coordinate accumulation of chloroplast and mitochondrial transcripts in the red alga Cyanidioschyzon merolae

The mitochondria and chloroplasts in plant cells are originated from bacterial endosymbioses, and they still replicate their own genome and divide in a similar manner as their ancestors did. It is thus likely that the organelle transcription is coordinated with its proliferation cycle. However, this possibility has not extensively been explored to date, because in most plant cells there are many mitochondria and chloroplasts that proliferate asynchronously. It is generally believed that the gene transfer from the organellar to nuclear genome has enabled nuclear control of the organelle functions during the evolution of eukaryotic plant cells. Nevertheless, no significant relationship has been reported between the organelle transcriptome and the host cell cycle even in Chlamydomonas reinhardtii. While the organelle proliferation cycle is not coordinated with the cell cycle in vascular plants, in the unicellular red alga Cyanidioschyzon merolae that contains only one mitochondrion, one chloroplast, and one nucleus per cell, each of the organelles is known to proliferate at a specific phase of the cell cycle. Here, we show that the expression of most of the organelle genes is highly coordinated with the cell cycle phases as well as with light regimes in clustering analyses. In addition, a strong correlation was observed between the gene expression profiles in the mitochondrion and chloroplast, resulting in the identification of a network of functionally related genes that are co-expressed during organelle proliferation.

Autotaxin and its product lysophosphatidic acid suppress brown adipose differentiation and promote diet-induced obesity in mice

Brown adipose tissue is a thermogenic organ that dissipates stored energy as heat to maintain body temperature. This process may also provide protection from development of diet-induced obesity. We report that the bioactive lipid mediator lysophosphatidic acid (LPA) markedly decreases differentiation of cultured primary brown adipocyte precursors, whereas potent selective inhibitors of the LPA-generating enzyme autotaxin (ATX) promote differentiation. Transgenic mice overexpressing ATX exhibit reduced expression of brown adipose tissue-related genes in peripheral white adipose tissue and accumulate significantly more fat than wild-type controls when fed a high-fat diet. Our results indicate that ATX and its product LPA are physiologically relevant negative regulators of brown fat adipogenesis and are consistent with a model in which a decrease in mature peripheral brown adipose tissue results in increased susceptibility to diet-induced obesity in mice.

Identification of single- and multiple-class specific signature genes from gene expression profiles by group marker index

Informative genes from microarray data can be used to construct prediction model and investigate biological mechanisms. Differentially expressed genes, the main targets of most gene selection methods, can be classified as single- and multiple-class specific signature genes. Here, we present a novel gene selection algorithm based on a Group Marker Index (GMI), which is intuitive, of low-computational complexity, and efficient in identification of both types of genes. Most gene selection methods identify only single-class specific signature genes and cannot identify multiple-class specific signature genes easily. Our algorithm can detect de novo certain conditions of multiple-class specificity of a gene and makes use of a novel non-parametric indicator to assess the discrimination ability between classes. Our method is effective even when the sample size is small as well as when the class sizes are significantly different. To compare the effectiveness and robustness we formulate an intuitive template-based method and use four well-known datasets. We demonstrate that our algorithm outperforms the template-based method in difficult cases with unbalanced distribution. Moreover, the multiple-class specific genes are good biomarkers and play important roles in biological pathways. Our literature survey supports that the proposed method identifies unique multiple-class specific marker genes (not reported earlier to be related to cancer) in the Central Nervous System data. It also discovers unique biomarkers indicating the intrinsic difference between subtypes of lung cancer. We also associate the pathway information with the multiple-class specific signature genes and cross-reference to published studies. We find that the identified genes participate in the pathways directly involved in cancer development in leukemia data. Our method gives a promising way to find genes that can involve in pathways of multiple diseases and hence opens up the possibility of using an existing drug on other diseases as well as designing a single drug for multiple diseases.

Differential gene expression in liver and small intestine from lactating rats compared to age-matched virgin controls detects increased mRNA of cholesterol biosynthetic genes

Background

Lactation increases energy demands four- to five-fold, leading to a two- to three-fold increase in food consumption, requiring a proportional adjustment in the ability of the lactating dam to absorb nutrients and to synthesize critical biomolecules, such as cholesterol, to meet the dietary needs of both the offspring and the dam. The size and hydrophobicity of the bile acid pool increases during lactation, implying an increased absorption and disposition of lipids, sterols, nutrients, and xenobiotics. In order to investigate changes at the transcriptomics level, we utilized an exon array and calculated expression levels to investigate changes in gene expression in the liver, duodenum, jejunum, and ileum of lactating dams when compared against age-matched virgin controls.

Results

A two-way mixed models ANOVA was applied to detect differentially expressed genes. Significance calls were defined as a p < 0.05 for the overall physiologic state effect (lactation vs. control), and a within tissue pairwise comparison of p < 0.01. The proportion of false positives, an estimate of the ratio of false positives in the list of differentially expressed genes, was calculated for each tissue. The number of differentially expressed genes was 420 in the liver, 337 in the duodenum, 402 in the jejunum, and 523 in the ileum. The list of differentially expressed genes was in turn analyzed by Ingenuity Pathways Analysis (IPA) to detect biological pathways that were overrepresented. In all tissues, sterol regulatory element binding protein (Srebp)-regulated genes involved in cholesterol synthesis showed increased mRNA expression, with the fewest changes detected in the jejunum. We detected increased Scap mRNA in the liver only, suggesting an explanation for the difference in response to lactation between the liver and small intestine. Expression of Cyp7a1, which catalyzes the rate limiting step in the bile acid biosynthetic pathway, was also significantly increased in liver. In addition, decreased levels of mRNA associated with T-cell signaling were found in the jejunum and ileum. Several members of the Solute Carrier (SLC) and Adenosine Triphosphate Binding Cassette (ABC) superfamilies of membrane transporters were found to be differentially expressed; these genes may play a role in differences in nutrient and xenobiotic absorption and disposition. mRNA expression of SLC39a4_predicted, a zinc transporter, was increased in all tissues, suggesting that it is involved in increased zinc uptake during lactation. Microarray data are available through GEO under GSE19175.

Conclusions

We detected differential expression of mRNA from several pathways in lactating dams, including upregulation of the cholesterol biosynthetic pathway in liver and intestine, consistent with Srebp activation. Differential T-Cell signaling in the two most distal regions of the small intestine (ileum and jejunum) was also noted, as well as differential expression of transporters that likely play a key role in nutrient uptake.

Analyzing microarray data with transitive directed acyclic graphs

Post hoc assignment of patterns determined by all pairwise comparisons in microarray experiments with multiple treatments has been proven to be useful in assessing treatment effects. We propose the usage of transitive directed acyclic graphs (tDAG) as the representation of these patterns and show that such representation can be useful in clustering treatment effects, annotating existing clustering methods, and analyzing sample sizes. Advantages of this approach include: (1) unique and descriptive meaning of each cluster in terms of how genes respond to all pairs of treatments; (2) insensitivity of the observed patterns to the number of genes analyzed; and (3) a combinatorial perspective to address the sample size problem by observing the rate of contractible tDAG as the number of replicates increases. The advantages and overall utility of the method in elaborating drug structure activity relationships are exemplified in a controlled study with real and simulated data.