Probing genetic algorithms for feature selection in comprehensive metabolic profiling approach

A mid-infrared spectroscopy-random forest system for the origin tracing of Chinese geographical indication Aconiti Lateralis Radix Praeparata

Reliable origin certification methods are essential for the protection of high-value genuine medicinal material with designated origins and geographical indication (GI) products. Aconiti Lateralis Radix Praeparata (Fuzi), one well-known traditional Chinese medicine and geographical indication products have remarkable efficacy and wide clinical application, with high demand in domestic and international markets. The efficacy and price of Fuzi from different origins vary, and it is difficult for the general public to accurately identify them through traditional experience. The mass spectrometry detection technology based on the plant metabolomics is tedious and lengthy in test sample preparation, complicated in operation, long in detection time, and low in reproducibility. As a sophisticated, green, fast, and low-loss detection technique, infrared spectroscopy is integrated by machine learning to bring new ways for quality regulation and control of traditional Chinese medicines. An analytical method based on mid-infrared spectroscopy combined with a random forest algorithm was developed to verify the geographical origin of authentic herbs and/or GI products. The method successfully predicted and classified three varieties of Chinese GI Fuzi and four varieties of non-GI Fuzi. In this study, an environment-friendly traceability strategy with fast analysis, low sample loss and high precision was used to provide a new strategy for identifying the origin of Fuzi.

An integrated spectroscopic strategy to trace the geographical origins of emblic medicines: Application for the quality assessment of natural medicines

Emblic medicine is a popular natural source in the world due to its outstanding healthcare and therapeutic functions. Our preliminary results indicated that the quality of emblic medicines might have an apparent regional variation. A rapid and effective geographical traceability system has not been designed yet. To trace the geographical origins so that their quality can be controlled, an integrated spectroscopic strategy including spectral pretreatment, outlier diagnosis, feature selection, data fusion, and machine learning algorithm was proposed. A featured data matrix (245 × 220) was successfully generated, and a carefully adjusted RF machine learning algorithm was utilized to develop the geographical traceability model. The results demonstrate that the proposed strategy is effective and can be generalized. Sensitivity (SEN), specificity (SPE) and accuracy (ACC) of 97.65%, 99.85% and 97.63% for the calibrated set, as well as 100.00% predictive efficiency, were obtained using this spectroscopic analysis strategy. Our study has created an integrated analysis process for multiple spectral data, which can achieve a rapid, nondestructive and green quality detection for emblic medicines originating from seventeen geographical origins.

Untargeted Metabolomic Profile for the Detection of Prostate Carcinoma-Preliminary Results from PARAFAC2 and PLS-DA Models

Prostate-specific antigen (PSA) is the main biomarker for the screening of prostate cancer (PCa), which has a high sensibility (higher than 80%) that is negatively offset by its poor specificity (only 30%, with the European cut-off of 4 ng/mL). This generates a large number of useless biopsies, involving both risks for the patients and costs for the national healthcare systems. Consequently, efforts were recently made to discover new biomarkers useful for PCa screening, including our proposal of interpreting a multi-parametric urinary steroidal profile with multivariate statistics. This approach has been expanded to investigate new alleged biomarkers by the application of untargeted urinary metabolomics. Urine samples from 91 patients (43 affected by PCa; 48 by benign hyperplasia) were deconjugated, extracted in both basic and acidic conditions, derivatized with different reagents, and analyzed with different gas chromatographic columns. Three-dimensional data were obtained from full-scan electron impact mass spectra. The PARADISe software, coupled with NIST libraries, was employed for the computation of PARAFAC2 models, the extraction of the significative components (alleged biomarkers), and the generation of a semiquantitative dataset. After variables selection, a partial least squares-discriminant analysis classification model was built, yielding promising performances. The selected biomarkers need further validation, possibly involving, yet again, a targeted approach.

Automated Comparative Metabolite Profiling of Large LC-ESIMS Data Sets in an ACD/MS Workbook Suite Add-in, and Data Clustering on a New Open-Source Web Platform FreeClust

The technological development of LC-MS instrumentation has led to significant improvements of performance and sensitivity, enabling high-throughput analysis of complex samples, such as plant extracts. Most software suites allow preprocessing of LC-MS chromatograms to obtain comprehensive information on single constituents. However, more advanced processing needs, such as the systematic and unbiased comparative metabolite profiling of large numbers of complex LC-MS chromatograms remains a challenge. Currently, users have to rely on different tools to perform such data analyses. We developed a two-step protocol comprising a comparative metabolite profiling tool integrated in ACD/MS Workbook Suite, and a web platform developed in R language designed for clustering and visualization of chromatographic data. Initially, all relevant chromatographic and spectroscopic data (retention time, molecular ions with the respective ion abundance, and sample names) are automatically extracted and assembled in an Excel spreadsheet. The file is then loaded into an online web application that includes various statistical algorithms and provides the user with tools to compare and visualize the results in intuitive 2D heatmaps. We applied this workflow to LC-ESIMS profiles obtained from 69 honey samples. Within few hours of calculation with a standard PC, honey samples were preprocessed and organized in clusters based on their metabolite profile similarities, thereby highlighting the common metabolite patterns and distributions among samples. Implementation in the ACD/Laboratories software package enables ulterior integration of other analytical data, and in silico prediction tools for modern drug discovery.

HILIC-MS for metabolomics: An attractive and complementary approach to RPLC-MS

Hydrophilic interaction chromatography (HILIC) is an emerging separation mode of liquid chromatography (LC). Using highly hydrophilic stationary phases capable of retaining polar/ionic metabolites, and accompany with high organic content mobile phase that offer readily compatibility with mass spectrometry (MS) has made HILIC an attractive complementary tool to the widely used reverse-phase (RP) chromatographic separations in metabolomic studies. The combination of HILIC and RPLC coupled with an MS detector expands the number of detected analytes and provides more comprehensive metabolite coverage than use of only RP chromatography. This review describes the recent applications of HILIC-MS/MS in metabolomic studies, ranging from amino acids, lipids, nucleotides, organic acids, pharmaceuticals, and metabolites of specific nature. The biological systems investigated include microbials, cultured cell line, plants, herbal medicine, urine, and serum as well as tissues from animals and humans. Owing to its unique capability to measure more-polar biomolecules, the HILIC separation technique would no doubt enhance the comprehensiveness of metabolite detection, and add significant value for metabolomic investigations. © 2014 Wiley Periodicals, Inc. Mass Spec Rev 35:574-600, 2016.

Microbiota-driven transcriptional changes in prefrontal cortex override genetic differences in social behavior

Gene-environment interactions impact the development of neuropsychiatric disorders, but the relative contributions are unclear. Here, we identify gut microbiota as sufficient to induce depressive-like behaviors in genetically distinct mouse strains. Daily gavage of vehicle (dH2O) in nonobese diabetic (NOD) mice induced a social avoidance behavior that was not observed in C57BL/6 mice. This was not observed in NOD animals with depleted microbiota via oral administration of antibiotics. Transfer of intestinal microbiota, including members of the Clostridiales, Lachnospiraceae and Ruminococcaceae, from vehicle-gavaged NOD donors to microbiota-depleted C57BL/6 recipients was sufficient to induce social avoidance and change gene expression and myelination in the prefrontal cortex. Metabolomic analysis identified increased cresol levels in these mice, and exposure of cultured oligodendrocytes to this metabolite prevented myelin gene expression and differentiation. Our results thus demonstrate that the gut microbiota modifies the synthesis of key metabolites affecting gene expression in the prefrontal cortex, thereby modulating social behavior.

DOI:http://dx.doi.org/10.7554/eLife.13442.001

A combination of genes and environmental factors underlie an individual’s risk of developing a mental illness. Among the environmental factors, it is becoming clear that communication between the gut and the brain is involved, but we do not understand how these two organs communicate. Our gut contains a variety of bacteria that help us to digest food and there is some evidence that changes in these bacterial communities can influence our mental health.

Transplanting feces from one individual to the gut of another is one way to alter the communities of bacteria in the gut. Here, Gacias et al. investigated whether fecal transplants are sufficient to induce social avoidance behavior – a symptom of depression – in mice. The experiments show that introducing specific combinations of bacteria into the gut is indeed able to cause healthy adult mice to avoid social interactions. This effect was caused by changes in the “myelin” sheath that surrounds many nerve fibers and could be prevented by giving the mice antibiotics, which decreased the number of bacteria in the gut.

Further experiments revealed that the mice that became depressed after fecal transplants had higher levels of a molecule called cresol, which is produced by certain gut bacteria. Gacias et al. found that cresol is able to reduce the amount of myelin produced by brain cells. Therefore, these findings show that changing the communities of bacteria in the gut can result in the accumulation of molecules that influence social behavior. Future work will aim to identify bacteria that can reduce the amount of cresol produced in the gut, which may have the potential to treat depression.

DOI:http://dx.doi.org/10.7554/eLife.13442.002

Metabolite changes in conifer buds and needles during forced bud break in Norway spruce (Picea abies) and European silver fir (Abies alba)

Environmental changes such as early spring and warm spells induce bud burst and photosynthetic processes in cold-acclimated coniferous trees and consequently, cellular metabolism in overwintering needles and buds. The purpose of the study was to examine metabolism in conifers under forced deacclimation (artificially induced spring) by exposing shoots of Picea abies (boreal species) and Abies alba (temperate species) to a greenhouse environment (22°C, 16/8 h D/N cycle) over a 9 weeks period. Each week, we scored bud opening and collected samples for GC/MS-based metabolite profiling. We detected a total of 169 assigned metabolites and 80 identified metabolites, comprising compounds such as mono- and disaccharides, Krebs cycle acids, amino acids, polyols, phenolics, and phosphorylated structures. Untargeted multivariate statistical analysis based on PCA and cluster analysis segregated samples by species, tissue type, and stage of tissue deacclimations. Similar patterns of metabolic regulation in both species were observed in buds (amino acids, Krebs cycle acids) and needles (hexoses, pentoses, and Krebs cycle acids). Based on correlation of bud opening score with compound levels, distinct metabolites could be associated with bud and shoot development, including amino acids, sugars, and acids with known osmolyte function, and secondary metabolites. This study has shed light on how elevated temperature affects metabolism in buds and needles of conifer species during the deacclimation phase, and contributes to the discussion about how phenological characters in conifers may respond to future global warming.

Ad-hoc blocked design for the robustness study in the determination of dichlobenil and 2,6-dichlorobenzamide in onions by programmed temperature vaporization-gas chromatography-mass spectrometry

An 'ad-hoc' experimental design to handle the robustness study for the simultaneous determination of dichlobenil and its main metabolite (2,6-dichlorobenzamide) in onions by programmed temperature vaporization-gas chromatography-mass spectrometry (PTV-GC-MS) is performed. Eighteen experimental factors were considered; 7 related with the extraction and clean up step, 8 with the PTV injection step and 3 factors related with the derivatization step. Therefore, a high number of experiments must be carried out that cannot be conducted in one experimental session and, as a consequence, the experiments of the robustness study must be performed in several sessions or blocks. The procedure to obtain an experimental design suitable for this task works by simultaneously minimizing the joint confidence region for the coefficient estimates and the correlation among them and with the block. In this way, the effect of the factors is not aliased with the block avoiding possible misinterpretations of the effects of the experimental factors on the analytical responses. The developed experimental design is coupled to the PARAFAC2 method, which allows solving some specific problems in chromatography when working with complex matrix such as co-elution of interferents (including silylation artifacts from the derivatization step) and small shifts in the retention time and, besides, the unequivocal identification of the target compounds according to document SANCO/12571/2013.

A comprehensive workflow of mass spectrometry-based untargeted metabolomics in cancer metabolic biomarker discovery using human plasma and urine

Current available biomarkers lack sensitivity and/or specificity for early detection of cancer. To address this challenge, a robust and complete workflow for metabolic profiling and data mining is described in details. Three independent and complementary analytical techniques for metabolic profiling are applied: hydrophilic interaction liquid chromatography (HILIC-LC), reversed-phase liquid chromatography (RP-LC), and gas chromatography (GC). All three techniques are coupled to a mass spectrometer (MS) in the full scan acquisition mode, and both unsupervised and supervised methods are used for data mining. The univariate and multivariate feature selection are used to determine subsets of potentially discriminative predictors. These predictors are further identified by obtaining accurate masses and isotopic ratios using selected ion monitoring (SIM) and data-dependent MS/MS and/or accurate mass MSn ion tree scans utilizing high resolution MS. A list combining all of the identified potential biomarkers generated from different platforms and algorithms is used for pathway analysis. Such a workflow combining comprehensive metabolic profiling and advanced data mining techniques may provide a powerful approach for metabolic pathway analysis and biomarker discovery in cancer research. Two case studies with previous published data are adapted and included in the context to elucidate the application of the workflow.

Direct and unbiased information recovery from liquid chromatography-mass spectrometry raw data for phenotype-differentiating metabolites based on screening window coefficient of ion currents

A reworking of a data mining strategy, in which statistical treatment of raw data from liquid chromatography-mass spectrometry (LC-MS) precedes recognition of chromatographic peaks, is presented. In this algorithm the tR-m/z plane of LC-MS data is divided into equal-sized segments of twelve seconds by one m/z unit each, and the total ion currents in corresponding segments as specified by the tR-m/z pair from multiple LC-MS runs are evaluated to generate mean ion currents (μ) and standard deviations (σ). The μ's and σ's of the segments, derived from contrasting classes of LC-MS data set (e.g., resistant-susceptible, case-control, etc.), are used to calculate the Z-factor (screening window coefficient) which is in turn used to rank the segments. Chromatographic peaks are recognized only where the ion currents are shown to differentiate the classes. The result-reporting format enables detection of positive as well as negative correlations between ion intensities and biological traits under study and thus points to the presence of potentially phenotype-discriminating metabolites. Examples of data analyses are presented, in which ions that may distinguish resistant and susceptible species of Aesculus to the leaf-miner Cameraria ohridella were detected.

Mechanism of gallic acid biosynthesis in bacteria (Escherichia coli) and walnut (Juglans regia)

Gallic acid (GA), a key intermediate in the synthesis of plant hydrolysable tannins, is also a primary anti-inflammatory, cardio-protective agent found in wine, tea, and cocoa. In this publication, we reveal the identity of a gene and encoded protein essential for GA synthesis. Although it has long been recognized that plants, bacteria, and fungi synthesize and accumulate GA, the pathway leading to its synthesis was largely unknown. Here we provide evidence that shikimate dehydrogenase (SDH), a shikimate pathway enzyme essential for aromatic amino acid synthesis, is also required for GA production. Escherichia coli (E. coli) aroE mutants lacking a functional SDH can be complemented with the plant enzyme such that they grew on media lacking aromatic amino acids and produced GA in vitro. Transgenic Nicotiana tabacum lines expressing a Juglans regia SDH exhibited a 500% increase in GA accumulation. The J. regia and E. coli SDH was purified via overexpression in E. coli and used to measure substrate and cofactor kinetics, following reduction of NADP(+) to NADPH. Reversed-phase liquid chromatography coupled to electrospray mass spectrometry (RP-LC/ESI-MS) was used to quantify and validate GA production through dehydrogenation of 3-dehydroshikimate (3-DHS) by purified E. coli and J. regia SDH when shikimic acid (SA) or 3-DHS were used as substrates and NADP(+) as cofactor. Finally, we show that purified E. coli and J. regia SDH produced GA in vitro.

Comprehensive mass spectrometry based metabolic profiling of blood plasma reveals potent discriminatory classifiers of pancreatic cancer

Poor outcome of pancreatic cancer necessitates development of an early diagnostic method to reduce mortality. No reliable early diagnostic test for pancreatic cancer detection has been developed and validated to date. In the current study, metabolic profiling of plasma samples from selected cancer patients and noncancerous controls was performed to seek novel metabolic biomarkers of pancreatic cancer. A comprehensive mass spectrometry based analytical platform established at the Metabolomics Core of the UC Davis Genome Center allowed detection of multiple compounds previously unreported in plasma from pancreatic cancer patients. It was found that selective amino acids, bile acids, and polar lipids were detected with increased or decreased levels in pancreatic cancer samples compared to controls. These findings on blood plasma levels of the relevant metabolites might be very useful clinical parameters for early diagnosis of pancreatic cancer.

Metabolomics tools for identifying biomarkers for neuropsychiatric diseases

The repertoire of biochemicals (or small molecules) present in cells, tissue, and body fluids is known as the metabolome. Today, clinicians utilize only a very small part of the information contained in the metabolome, as revealed by the quantification of a limited set of analytes to gain information on human health. Examples include measuring glucose or cholesterol to monitor diabetes and cardiovascular health, respectively. With a focus on comprehensively studying the metabolome, the rapidly growing field of metabolomics captures the metabolic state of organisms at the global or "-omics" level. Given that the overall health status of an individual is captured by his or her metabolic state, which is a reflection of what has been encoded by the genome and modified by environmental factors, metabolomics has the potential to have a great impact upon medical practice by providing a wealth of relevant biochemical data. Metabolomics promises to improve current, single metabolites-based clinical assessments by identifying metabolic signatures (biomarkers) that embody global biochemical changes in disease, predict responses to treatment or medication side effects (pharmachometabolomics). State of the art metabolomic analytical platforms and informatics tools are being used to map potential biomarkers for a multitude of disorders including those of the central nervous system (CNS). Indeed, CNS disorders are linked to disturbances in metabolic pathways related to neurotransmitter systems (dopamine, serotonin, GABA and glutamate); fatty acids such as arachidonic acid-cascade; oxidative stress and mitochondrial function. Metabolomics tools are enabling us to map in greater detail perturbations in many biochemical pathways and links among these pathways this information is key for development of biomarkers that are disease-specific. In this review, we elaborate on some of the concepts and technologies used in metabolomics and its promise for biomarker discovery. We also highlight early findings from metabolomic studies in CNS disorders such as schizophrenia, Major Depressive Disorder (MDD), Bipolar Disorder (BD), Amyotrophic lateral sclerosis (ALS) and Parkinson's disease (PD).