Joachim Selbig

Extension of the visualization tool mapman to allow statistical analysis of arrays, display of coresponding genes, and comparison with known responses

MapMan is a user-driven tool that displays large genomics datasets onto diagrams of metabolic pathways or other processes. Here, we present new developments, including improvements of the gene assignments and the user interface, a strategy to visualize multilayered datasets, the incorporation of statistics packages, and extensions of the software to incorporate more biological information including visualization of coresponding genes and horizontal searches for similar global responses across large numbers of arrays.

A Robot-Based Platform to Measure Multiple Enzyme Activities in Arabidopsis Using a Set of Cycling Assays: Comparison of Changes of Enzyme Activities and Transcript Levels during Diurnal Cycles and in Prolonged Darkness

A platform has been developed to measure the activity of 23 enzymes that are involved in central carbon and nitrogen metabolism in Arabidopsis thaliana. Activities are assayed in optimized stopped assays and the product then determined using a suite of enzyme cycling assays. The platform requires inexpensive equipment, is organized in a modular manner to optimize logistics, calculates results automatically, combines high sensitivity with throughput, can be robotized, and has a throughput of three to four activities in 100 samples per person/day. Several of the assays, including those for sucrose phosphate synthase, ADP glucose pyrophosphorylase (AGPase), ferredoxin-dependent glutamate synthase, glycerokinase, and shikimate dehydrogenase, provide large advantages over previous approaches. This platform was used to analyze the diurnal changes of enzyme activities in wild-type Columbia-0 (Col-0) and the starchless plastid phosphoglucomutase (pgm) mutant, and in Col-0 during a prolongation of the night. The changes of enzyme activities were compared with the changes of transcript levels determined with the Affymetrix ATH1 array. Changes of transcript levels typically led to strongly damped changes of enzyme activity. There was no relation between the amplitudes of the diurnal changes of transcript and enzyme activity. The largest diurnal changes in activity were found for AGPase and nitrate reductase. Examination of the data and comparison with the literature indicated that these are mainly because of posttranslational regulation. The changes of enzyme activity are also strongly delayed, with the delay varying from enzyme to enzyme. It is proposed that enzyme activities provide a quasi-stable integration of regulation at several levels and provide useful data for the characterization and diagnosis of different physiological states. As an illustration, a decision tree constructed using data from Col-0 during diurnal changes and a prolonged dark treatment was used to show that, irrespective of the time of harvest during the diurnal cycle, the pgm mutant resembles a wild-type plant that has been exposed to a 3 d prolongation of the night.

pcaMethods—a bioconductor package providing PCA methods for incomplete data

UNLABELLED pcaMethods is a Bioconductor compliant library for computing principal component analysis (PCA) on incomplete data sets. The results can be analyzed directly or used to estimate missing values to enable the use of missing value sensitive statistical methods. The package was mainly developed with microarray and metabolite data sets in mind, but can be applied to any other incomplete data set as well. AVAILABILITY http://www.bioconductor.org

Starch as a major integrator in the regulation of plant growth

Rising demand for food and bioenergy makes it imperative to breed for increased crop yield. Vegetative plant growth could be driven by resource acquisition or developmental programs. Metabolite profiling in 94 Arabidopsis accessions revealed that biomass correlates negatively with many metabolites, especially starch. Starch accumulates in the light and is degraded at night to provide a sustained supply of carbon for growth. Multivariate analysis revealed that starch is an integrator of the overall metabolic response. We hypothesized that this reflects variation in a regulatory network that balances growth with the carbon supply. Transcript profiling in 21 accessions revealed coordinated changes of transcripts of more than 70 carbon-regulated genes and identified 2 genes (myo-inositol-1-phosphate synthase, a Kelch-domain protein) whose transcripts correlate with biomass. The impact of allelic variation at these 2 loci was shown by association mapping, identifying them as candidate lead genes with the potential to increase biomass production.

The metabolic signature related to high plant growth rate in Arabidopsis thaliana.

The decline of available fossil fuel reserves has triggered world-wide efforts to develop alternative energy sources based on plant biomass. Detailed knowledge of the relations of metabolism and biomass accumulation can be expected to yield powerful novel tools to accelerate and enhance energy plant breeding programs. We used metabolic profiling in the model Arabidopsis to study the relation between biomass and metabolic composition using a recombinant inbred line (RIL) population. A highly significant canonical correlation (0.73) was observed, revealing a close link between biomass and a specific combination of metabolites. Dividing the entire data set into training and test sets resulted in a median correlation between predicted and true biomass of 0.58. The demonstrated high predictive power of metabolic composition for biomass features this composite measure as an excellent biomarker and opens new opportunities to enhance plant breeding specifically in the context of renewable resources.

Parallel analysis of transcript and metabolic profiles: a new approach in systems biology

The past few years in the medical and biological sciences have been characterized by the advent of systems biology. However, despite the well‐known connectivity between the molecules described by transcriptomic, proteomic and metabolomic approaches, few studies have tried to correlate parameters across the various levels of systemic description. When comparing the discriminatory power of metabolic and RNA profiling to distinguish between different potato tuber systems, using the techniques described here suggests that metabolic profiling has a higher resolution than expression profiling. When applying pairwise transcript–metabolite correlation analyses, 571 of the 26,616 possible pairs showed significant correlation, most of which was novel and included several strong correlations to nutritionally important metabolites. We believe this approach to be of high potential value in the identification of candidate genes for modifying the metabolite content of biological systems.

Metabolomic and transcriptomic stress response of Escherichia coli.

Environmental fluctuations lead to a rapid adjustment of the physiology of Escherichia coli, necessitating changes on every level of the underlying cellular and molecular network. Thus far, the majority of global analyses of E. coli stress responses have been limited to just one level, gene expression. Here, we incorporate the metabolite composition together with gene expression data to provide a more comprehensive insight on system level stress adjustments by describing detailed time‐resolved E. coli response to five different perturbations (cold, heat, oxidative stress, lactose diauxie, and stationary phase). The metabolite response is more specific as compared with the general response observed on the transcript level and is reflected by much higher specificity during the early stress adaptation phase and when comparing the stationary phase response to other perturbations. Despite these differences, the response on both levels still follows the same dynamics and general strategy of energy conservation as reflected by rapid decrease of central carbon metabolism intermediates coinciding with downregulation of genes related to cell growth. Application of co‐clustering and canonical correlation analysis on combined metabolite and transcript data identified a number of significant condition‐dependent associations between metabolites and transcripts. The results confirm and extend existing models about co‐regulation between gene expression and metabolites demonstrating the power of integrated systems oriented analysis.

PhosPhAt: a database of phosphorylation sites in Arabidopsis thaliana and a plant-specific phosphorylation site predictor

The PhosPhAt database provides a resource consolidating our current knowledge of mass spectrometry-based identified phosphorylation sites in Arabidopsis and combines it with phosphorylation site prediction specifically trained on experimentally identified Arabidopsis phosphorylation motifs. The database currently contains 1187 unique tryptic peptide sequences encompassing 1053 Arabidopsis proteins. Among the characterized phosphorylation sites, there are over 1000 with unambiguous site assignments, and nearly 500 for which the precise phosphorylation site could not be determined. The database is searchable by protein accession number, physical peptide characteristics, as well as by experimental conditions (tissue sampled, phosphopeptide enrichment method). For each protein, a phosphorylation site overview is presented in tabular form with detailed information on each identified phosphopeptide. We have utilized a set of 802 experimentally validated serine phosphorylation sites to develop a method for prediction of serine phosphorylation (pSer) in Arabidopsis. An analysis of the current annotated Arabidopsis proteome yielded in 27 782 predicted phosphoserine sites distributed across 17 035 proteins. These prediction results are summarized graphically in the database together with the experimental phosphorylation sites in a whole sequence context. The Arabidopsis Protein Phosphorylation Site Database (PhosPhAt) provides a valuable resource to the plant science community and can be accessed through the following link http://phosphat.mpimp-golm.mpg.de

Metabolite fingerprinting: detecting biological features by independent component analysis

MOTIVATION Metabolite fingerprinting is a technology for providing information from spectra of total compositions of metabolites. Here, spectra acquisitions by microchip-based nanoflow-direct-infusion QTOF mass spectrometry, a simple and high throughput technique, is tested for its informative power. As a simple test case we are using Arabidopsis thaliana crosses. The question is how metabolite fingerprinting reflects the biological background. In many applications the classical principal component analysis (PCA) is used for detecting relevant information. Here a modern alternative is introduced-the independent component analysis (ICA). Due to its independence condition, ICA is more suitable for our questions than PCA. However, ICA has not been developed for a small number of high-dimensional samples, therefore a strategy is needed to overcome this limitation. RESULTS To apply ICA successfully it is essential first to reduce the high dimension of the dataset, by using PCA. The number of principal components determines the quality of ICA significantly, therefore we propose a criterion for estimating the optimal dimension automatically. The kurtosis measure is used to order the extracted components to our interest. Applied to our A. thaliana data, ICA detects three relevant factors, two biological and one technical, and clearly outperforms the PCA.

Diversity and complexity of HIV-1 drug resistance: a bioinformatics approach to predicting phenotype from genotype.

Drug resistance testing has been shown to be beneficial for clinical management of HIV type 1 infected patients. Whereas phenotypic assays directly measure drug resistance, the commonly used genotypic assays provide only indirect evidence of drug resistance, the major challenge being the interpretation of the sequence information. We analyzed the significance of sequence variations in the protease and reverse transcriptase genes for drug resistance and derived models that predict phenotypic resistance from genotypes. For 14 antiretroviral drugs, both genotypic and phenotypic resistance data from 471 clinical isolates were analyzed with a machine learning approach. Information profiles were obtained that quantify the statistical significance of each sequence position for drug resistance. For the different drugs, patterns of varying complexity were observed, including between one and nine sequence positions with substantial information content. Based on these information profiles, decision tree classifiers were generated to identify genotypic patterns characteristic of resistance or susceptibility to the different drugs. We obtained concise and easily interpretable models to predict drug resistance from sequence information. The prediction quality of the models was assessed in leave-one-out experiments in terms of the prediction error. We found prediction errors of 9.6–15.5% for all drugs except for zalcitabine, didanosine, and stavudine, with prediction errors between 25.4% and 32.0%. A prediction service is freely available at http://cartan.gmd.de/geno2pheno.html.