Signal Peptide Efficiency: From High-Throughput Data to Prediction and Explanation.

The passage of proteins across biological membranes via the general secretory (Sec) pathway is a universally conserved process with critical functions in cell physiology and important industrial applications. Proteins are directed into the Sec pathway by a signal peptide at their N-terminus. Estimating the impact of physicochemical signal peptide features on protein secretion levels has not been achieved so far, partially due to the extreme sequence variability of signal peptides. To elucidate relevant features of the signal peptide sequence that influence secretion efficiency, an evaluation of ∼12,000 different designed signal peptides was performed using a novel miniaturized high-throughput assay. The results were used to train a machine learning model, and a post-hoc explanation of the model is provided. By describing each signal peptide with a selection of 156 physicochemical features, it is now possible to both quantify feature importance and predict the protein secretion levels directed by each signal peptide. Our analyses allow the detection and explanation of the relevant signal peptide features influencing the efficiency of protein secretion, generating a versatile tool for the de novo design and in silico evaluation of signal peptides.

Flavor Profiling Using Comprehensive Mass Spectrometry Analysis of Metabolites in Tomato Soups.

Trained sensory panels are regularly used to rate food products but do not allow for data-driven approaches to steer food product development. This study evaluated the potential of a molecular-based strategy by analyzing 27 tomato soups that were enhanced with yeast-derived flavor products using a sensory panel as well as LC-MS and GC-MS profiling. These data sets were used to build prediction models for 26 different sensory attributes using partial least squares analysis. We found driving separation factors between the tomato soups and metabolites predicting different flavors. Many metabolites were putatively identified as dipeptides and sulfur-containing modified amino acids, which are scientifically described as related to umami or having "garlic-like" and "onion-like" attributes. Proposed identities of high-impact sensory markers (methionyl-proline and asparagine-leucine) were verified using MS/MS. The overall results highlighted the strength of combining sensory data and metabolomics platforms to find new information related to flavor perception in a complex food matrix.

Corruption of the Pearson correlation coefficient by measurement error and its estimation, bias, and correction under different error models.

Correlation coefficients are abundantly used in the life sciences. Their use can be limited to simple exploratory analysis or to construct association networks for visualization but they are also basic ingredients for sophisticated multivariate data analysis methods. It is therefore important to have reliable estimates for correlation coefficients. In modern life sciences, comprehensive measurement techniques are used to measure metabolites, proteins, gene-expressions and other types of data. All these measurement techniques have errors. Whereas in the old days, with simple measurements, the errors were also simple, that is not the case anymore. Errors are heterogeneous, non-constant and not independent. This hampers the quality of the estimated correlation coefficients seriously. We will discuss the different types of errors as present in modern comprehensive life science data and show with theory, simulations and real-life data how these affect the correlation coefficients. We will briefly discuss ways to improve the estimation of such coefficients.

Endometriosis is associated with aberrant metabolite profiles in plasma.

OBJECTIVE: To identify metabolites that are associated with and predict the presence of endometriosis. DESIGN: Metabolomics study using state-of-the-art mass spectrometry approaches. SETTING: University hospital and universities. PATIENT(S): Twenty-five women with laparoscopically confirmed endometriosis (cases) and 19 women with laparoscopically documented absence of endometriosis (controls). None of the women included in this study had received oral contraception or GnRH agonists for a minimum of 1 month before blood collection. INTERVENTION(S): Plasma collection. MAIN OUTCOME MEASURE(S): Metabolite profiles were generated and interrogated using multiple mass spectrometry methods, that is, high performance liquid chromatography coupled with negative mode electrospray ionization tandem mass spectrometry, UPLC-MS/MS, and ultra performance liquid chromatography-electroSpray ionization-quadrupole time-of-flight (UPLC-ESI-Q-TOF). Metabolite groups investigated included phospholipids, glycerophospholipids, ether-phospholipids, cholesterol-esters, triacylglycerol, sphingolipids, free fatty acids, steroids, eicosanoids, and acylcarnitines. RESULT(S): A panel of acylcarnitines predicted the presence of endometriosis with 88.9% specificity and 81.5% sensitivity in human plasma, with a positive predictive value of 75%. However, due to data limitations the outcome of the receiver operating characteristic curve analysis was not significant. CONCLUSION(S): A diagnostic model based on acylcarnitines has the potential to predict the presence and stage of endometriosis.

Oxylipid Profile of Low-Dose Aspirin Exposure: A Pharmacometabolomics Study.

BACKGROUND: While aspirin is a well-established and generally effective anti-platelet agent, considerable inter-individual variation in drug response exists, for which mechanisms are not completely understood. Metabolomics allows for extensive measurement of small molecules in biological samples, enabling detailed mapping of pathways involved in drug response. METHODS AND RESULTS: We used a mass-spectrometry-based metabolomics platform to investigate the changes in the serum oxylipid metabolome induced by an aspirin intervention (14 days, 81 mg/day) in healthy subjects (n=156). We observed a global decrease in serum oxylipids in response to aspirin (25 metabolites decreased out of 30 measured) regardless of sex. This decrease was concomitant with a significant decrease in serum linoleic acid levels (-19%, P=1.3×10(-5)), one of the main precursors for oxylipid synthesis. Interestingly, several linoleic acid-derived oxylipids were not significantly associated with arachidonic-induced ex vivo platelet aggregation, a widely accepted marker of aspirin response, but were significantly correlated with platelet reactivity in response to collagen. CONCLUSIONS: Together, these results suggest that linoleic acid-derived oxylipids may contribute to the non-COX1 mediated variability in response to aspirin. Pharmacometabolomics allowed for more comprehensive interrogation of mechanisms of action of low dose aspirin and of variation in aspirin response.

MetDFBA: incorporating time-resolved metabolomics measurements into dynamic flux balance analysis.

Understanding cellular adaptation to environmental changes is one of the major challenges in systems biology. To understand how cellular systems react towards perturbations of their steady state, the metabolic dynamics have to be described. Dynamic properties can be studied with kinetic models but development of such models is hampered by limited in vivo information, especially kinetic parameters. Therefore, there is a need for mathematical frameworks that use a minimal amount of kinetic information. One of these frameworks is dynamic flux balance analysis (DFBA), a method based on the assumption that cellular metabolism has evolved towards optimal changes to perturbations. However, DFBA has some limitations. It is less suitable for larger systems because of the high number of parameters to estimate and the computational complexity. In this paper, we propose MetDFBA, a modification of DFBA, that incorporates measured time series of both intracellular and extracellular metabolite concentrations, in order to reduce both the number of parameters to estimate and the computational complexity. MetDFBA can be used to estimate dynamic flux profiles and, in addition, test hypotheses about metabolic regulation. In a first case study, we demonstrate the validity of our method by comparing our results to flux estimations based on dynamic 13C MFA measurements, which we considered as experimental reference. For these estimations time-resolved metabolomics data from a feast-famine experiment with Penicillium chrysogenum was used. In a second case study, we used time-resolved metabolomics data from glucose pulse experiments during aerobic growth of Saccharomyces cerevisiae to test various metabolic objectives.

Correlated measurement error hampers association network inference.

Modern chromatography-based metabolomics measurements generate large amounts of data in the form of abundances of metabolites. An increasingly popular way of representing and analyzing such data is by means of association networks. Ideally, such a network can be interpreted in terms of the underlying biology. A property of chromatography-based metabolomics data is that the measurement error structure is complex: apart from the usual (random) instrumental error there is also correlated measurement error. This is intrinsic to the way the samples are prepared and the analyses are performed and cannot be avoided. The impact of correlated measurement errors on (partial) correlation networks can be large and is not always predictable. The interplay between relative amounts of uncorrelated measurement error, correlated measurement error and biological variation defines this impact. Using chromatography-based time-resolved lipidomics data obtained from a human intervention study we show how partial correlation based association networks are influenced by correlated measurement error. We show how the effect of correlated measurement error on partial correlations is different for direct and indirect associations. For direct associations the correlated measurement error usually has no negative effect on the results, while for indirect associations, depending on the relative size of the correlated measurement error, results can become unreliable. The aim of this paper is to generate awareness of the existence of correlated measurement errors and their influence on association networks. Time series lipidomics data is used for this purpose, as it makes it possible to visually distinguish the correlated measurement error from a biological response. Underestimating the phenomenon of correlated measurement error will result in the suggestion of biologically meaningful results that in reality rest solely on complicated error structures. Using proper experimental designs that allow for the quantification of the size of correlated and uncorrelated errors, can help to identify suspicious connections in association networks constructed from (partial) correlations.

Network identification of hormonal regulation.

Relations among hormone serum concentrations are complex and depend on various factors, including gender, age, body mass index, diurnal rhythms and secretion stochastics. Therefore, endocrine deviations from healthy homeostasis are not easily detected or understood. A generic method is presented for detecting regulatory relations between hormones. This is demonstrated with a cohort of obese women, who underwent blood sampling at 10 minute intervals for 24-hours. The cohort was treated with bromocriptine in an attempt to clarify how hormone relations change by treatment. The detected regulatory relations are summarized in a network graph and treatment-induced changes in the relations are determined. The proposed method identifies many relations, including well-known ones. Ultimately, the method provides ways to improve the description and understanding of normal hormonal relations and deviations caused by disease or treatment.

A new approach to untargeted integration of high resolution liquid chromatography-mass spectrometry data.

Because of its high sensitivity and specificity, hyphenated mass spectrometry has become the predominant method to detect and quantify metabolites present in bio-samples relevant for all sorts of life science studies being executed. In contrast to targeted methods that are dedicated to specific features, global profiling acquisition methods allow new unspecific metabolites to be analyzed. The challenge with these so-called untargeted methods is the proper and automated extraction and integration of features that could be of relevance. We propose a new algorithm that enables untargeted integration of samples that are measured with high resolution liquid chromatography-mass spectrometry (LC-MS). In contrast to other approaches limited user interaction is needed allowing also less experienced users to integrate their data. The large amount of single features that are found within a sample is combined to a smaller list of, compound-related, grouped feature-sets representative for that sample. These feature-sets allow for easier interpretation and identification and as important, easier matching over samples. We show that the automatic obtained integration results for a set of known target metabolites match those generated with vendor software but that at least 10 times more feature-sets are extracted as well. We demonstrate our approach using high resolution LC-MS data acquired for 128 samples on a lipidomics platform. The data was also processed in a targeted manner (with a combination of automatic and manual integration) using vendor software for a set of 174 targets. As our untargeted extraction procedure is run per sample and per mass trace the implementation of it is scalable. Because of the generic approach, we envision that this data extraction lipids method will be used in a targeted as well as untargeted analysis of many different kinds of TOF-MS data, even CE- and GC-MS data or MRM. The Matlab package is available for download on request and efforts are directed toward a user-friendly Windows executable.

Pre-processing liquid chromatography/high-resolution mass spectrometry data: extracting pure mass spectra by deconvolution from the invariance of isotopic distribution.

RATIONALE: Mass spectra obtained by deconvolution of liquid chromatography/high-resolution mass spectrometry (LC/HRMS) data can be impaired by non-informative mass-over-charge (m/z) channels. This impairment of mass spectra can have significant negative influence on further post-processing, like quantification and identification. METHODS: A metric derived from the knowledge of errors in isotopic distribution patterns, and quality of the signal within a pre-defined mass chromatogram block, has been developed to pre-select all informative m/z channels. RESULTS: This procedure results in the clean-up of deconvoluted mass spectra by maintaining the intensity counts from m/z channels that originate from a specific compound/molecular ion, for example, molecular ion, adducts, (13) C-isotopes, multiply charged ions and removing all m/z channels that are not related to the specific peak. The methodology has been successfully demonstrated for two sets of high-resolution LC/MS data. CONCLUSIONS: The approach described is therefore thought to be a useful tool in the automatic processing of LC/HRMS data. It clearly shows the advantages compared to other approaches like peak picking and de-isotoping in the sense that all information is retained while non-informative data is removed automatically.

Quantification of in vivo oxidative damage in Caenorhabditis elegans during aging by endogenous F3-isoprostane measurement.

Oxidative damage is thought to be a major cause in development of pathologies and aging. However, quantification of oxidative damage is methodologically difficult. Here, we present a robust liquid chromatography-tandem mass spectrometry (LC-MS/MS) approach for accurate, sensitive, and linear in vivo quantification of endogenous oxidative damage in the nematode Caenorhabditis elegans, based on F3-isoprostanes. F3-isoprostanes are prostaglandin-like markers of oxidative damage derived from lipid peroxidation by Reactive Oxygen Species (ROS). Oxidative damage was quantified in whole animals and in multiple cellular compartments, including mitochondria and peroxisomes. Mutants of the mitochondrial electron transport proteins mev-1 and clk-1 showed increased oxidative damage levels. Furthermore, analysis of Superoxide Dismutase (sod) and Catalase (ctl) mutants uncovered that oxidative damage levels cannot be inferred from the phenotype of resistance to pro-oxidants alone and revealed high oxidative damage in a small group of chemosensory neurons. Longitudinal analysis of aging nematodes revealed that oxidative damage increased specifically with postreproductive age. Remarkably, aging of the stress-resistant and long-lived daf-2 insulin/IGF-1 receptor mutant involved distinct daf-16-dependent phases of oxidative damage including a temporal increase at young adulthood. These observations are consistent with a hormetic response to ROS.

A lipidomic analysis approach to evaluate the response to cholesterol-lowering food intake.

Plant sterols (PS) are well known to reduce serum levels of total cholesterol and LDL-cholesterol. Lipidomics potentially provides detailed information on a wide range of individual serum lipid metabolites, which may further add to our understanding of the biological effects of PS. In this study, lipidomics analysis was applied to serum samples from a placebo-controlled, parallel human intervention study (n = 97) of 4-week consumption of two PS-enriched, yoghurt drinks differing in fat content (based on 0.1% vs. 1.5% dairy fat). A comprehensive data analysis strategy was developed and implemented to assess and compare effects of two different PS-treatments and placebo treatment. The combination of univariate and multivariate data analysis approaches allowed to show significant effects of PS intake on the serum lipidome, and helped to distinguish them from fat content and non-specific effects. The PS-enriched 0.1% dairy fat yoghurt drink had a stronger impact on the lipidome than the 1.5% dairy fat yoghurt drink, despite similar LDL-cholesterol lowering effects. The PS-enriched 0.1% dairy fat yoghurt drink reduced levels of several sphingomyelins which correlated well with the reduction in LDL-cholesterol and can be explained by co-localization of sphingomyelins and cholesterol on the surface of LDL lipoprotein. Statistically significant reductions in serum levels of two lysophosphatidylcholines (LPC(16:1), LPC(20:1)) and cholesteryl arachidonate may suggest reduced inflammation and atherogenic potential. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1007/s11306-011-0384-2) contains supplementary material, which is available to authorized users.

Instrument and process independent binning and baseline correction methods for liquid chromatography-high resolution-mass spectrometry deconvolution.

Setting appropriate bin sizes to aggregate hyphenated high-resolution mass spectrometry data, belonging to similar mass over charge (m/z) channels, is vital to metabolite quantification and further identification. In a high-resolution mass spectrometer when mass accuracy (ppm) varies as a function of molecular mass, which usually is the case while reading m/z from low to high values, it becomes a challenge to determine suitable bin sizes satisfying all m/z ranges. Similarly, the chromatographic process within a hyphenated system, like any other controlled processes, introduces some process driven systematic behavior that ultimately distorts the mass chromatogram signal. This is especially seen in liquid chromatogram-mass spectrometry (LC-MS) measurements where the gradient of the solvent and the washing step cycle-part of the chromatographic process, produce a mass chromatogram with a non-uniform baseline along the retention time axis. Hence prior to any automatic signal decomposition techniques like deconvolution, it is a equally vital to perform the baseline correction step for absolute metabolite quantification. This paper will discuss an instrument and process independent solution to the binning and the baseline correction problem discussed above, seen together, as an effective pre-processing step toward liquid chromatography-high resolution-mass spectrometry (LC-HR-MS) data deconvolution.

Inferring differences in the distribution of reaction rates across conditions.

Elucidating changes in the distribution of reaction rates in metabolic pathways under different conditions is a central challenge in systems biology. Here we present a method for inferring regulation mechanisms responsible for changes in the distribution of reaction rates across conditions from correlations in time-resolved data. A reversal of correlations between conditions reveals information about regulation mechanisms. With the use of a small in silico hypothetical network, based on only the topology and directionality of a known pathway, several regulation scenarios can be formulated. Confronting these scenarios with experimental data results in a short list of possible pathway regulation mechanisms associated with the reversal of correlations between conditions. This procedure allows for the formulation of regulation scenarios without detailed prior knowledge of kinetics and for the inference of reaction rate changes without rate information. The method was applied to experimental time-resolved metabolomics data from multiple short-term perturbation-response experiments in S. cerevisiae across aerobic and anaerobic conditions. The method's output was validated against a detailed kinetic model of glycolysis in S. cerevisiae, which showed that the method can indeed infer the correct regulation scenario.

Detecting regulatory mechanisms in endocrine time series measurements.

The regulatory mechanisms underlying pulsatile secretion are complex, especially as it is partly controlled by other hormones and the combined action of multiple agents. Regulatory relations between hormones are not directly observable but may be deduced from time series measurements of plasma hormone concentrations. Variation in plasma hormone levels are the resultant of secretion and clearance from the circulation. A strategy is proposed to extract inhibition, activation, thresholds and circadian synchronicity from concentration data, using particular association methods. Time delayed associations between hormone concentrations and/or extracted secretion pulse profiles reveal the information on regulatory mechanisms. The above mentioned regulatory mechanisms are illustrated with simulated data. Additionally, data from a lean cohort of healthy control subjects is used to illustrate activation (ACTH and cortisol) and circadian synchronicity (ACTH and TSH) in real data. The simulation and the real data both consist of 145 equidistant samples per individual, matching a 24-hr time span with 10 minute intervals. The results of the simulation and the real data are in concordance.

Global test for metabolic pathway differences between conditions.

In many metabolomics applications there is a need to compare metabolite levels between different conditions, e.g., case versus control. There exist many statistical methods to perform such comparisons but only few of these explicitly take into account the fact that metabolites are connected in pathways or modules. Such a priori information on pathway structure can alleviate problems in, e.g., testing on individual metabolite level. In gene-expression analysis, Goeman's global test is used to this extent to determine whether a group of genes has a different expression pattern under changed conditions. We examined if this test can be generalized to metabolomics data. The goal is to determine if the behavior of a group of metabolites, belonging to the same pathway, is significantly related to a particular outcome of interest, e.g., case/control or environmental conditions. The results show that the global test can indeed be used in such situations. This is illustrated with extensive intracellular metabolomics data from Escherichia coli and Saccharomyces cerevisiae under different environmental conditions.

Generic framework for high-dimensional fixed-effects ANOVA.

In functional genomics it is more rule than exception that experimental designs are used to generate the data. The samples of the resulting data sets are thus organized according to this design and for each sample many biochemical compounds are measured, e.g. typically thousands of gene-expressions or hundreds of metabolites. This results in high-dimensional data sets with an underlying experimental design. Several methods have recently become available for analyzing such data while utilizing the underlying design. We review these methods by putting them in a unifying and general framework to facilitate understanding the (dis-)similarities between the methods. The biological question dictates which method to use and the framework allows for building new methods to accommodate a range of such biological questions. The framework is built on well known fixed-effect ANOVA models and subsequent dimension reduction. We present the framework both in matrix algebra as well as in more insightful geometrical terms. We show the workings of the different special cases of our framework with a real-life metabolomics example from nutritional research and a gene-expression example from the field of virology.

Simplivariate models: uncovering the underlying biology in functional genomics data.

One of the first steps in analyzing high-dimensional functional genomics data is an exploratory analysis of such data. Cluster Analysis and Principal Component Analysis are then usually the method of choice. Despite their versatility they also have a severe drawback: they do not always generate simple and interpretable solutions. On the basis of the observation that functional genomics data often contain both informative and non-informative variation, we propose a method that finds sets of variables containing informative variation. This informative variation is subsequently expressed in easily interpretable simplivariate components.We present a new implementation of the recently introduced simplivariate models. In this implementation, the informative variation is described by multiplicative models that can adequately represent the relations between functional genomics data. Both a simulated and two real-life metabolomics data sets show good performance of the method.

Reverse engineering of metabolic networks, a critical assessment.

Inferring metabolic networks from metabolite concentration data is a central topic in systems biology. Mathematical techniques to extract information about the network from data have been proposed in the literature. This paper presents a critical assessment of the feasibility of reverse engineering of metabolic networks, illustrated with a selection of methods. Appropriate data are simulated to study the performance of four representative methods. An overview of sampling and measurement methods currently in use for generating time-resolved metabolomics data is given and contrasted with the needs of the discussed reverse engineering methods. The results of this assessment show that if full inference of a real-world metabolic network is the goal there is a large discrepancy between the requirements of reverse engineering of metabolic networks and contemporary measurement practice. Recommendations for improved time-resolved experimental designs are given.