Machine Learning Predicting Atrial Fibrillation as an Adverse Event in the Warfarin and Aspirin in Reduced Cardiac Ejection Fraction (WARCEF) Trial.

BACKGROUND: Atrial fibrillation and heart failure commonly coexist due to shared pathophysiological mechanisms. Prompt identification of patients with heart failure at risk of developing atrial fibrillation would allow clinicians the opportunity to implement appropriate monitoring strategy and timely treatment, reducing the impact of atrial fibrillation on patients' health. METHODS: Four machine learning models combined with logistic regression and cluster analysis were applied post hoc to patient-level data from the Warfarin and Aspirin in Patients with Heart Failure and Sinus Rhythm (WARCEF) trial to identify factors that predict development of atrial fibrillation in patients with heart failure. RESULTS: Logistic regression showed that White divorced patients have a 1.75-fold higher risk of atrial fibrillation than White patients reporting other marital statuses. By contrast, similar analysis suggests that non-White patients who live alone have a 2.58-fold higher risk than those not living alone. Machine learning analysis also identified "marital status" and "live alone" as relevant predictors of atrial fibrillation. Apart from previously well-recognized factors, the machine learning algorithms and cluster analysis identified 2 distinct clusters, namely White and non-White ethnicities. This should serve as a reminder of the impact of social factors on health. CONCLUSION: The use of machine learning can prove useful in identifying novel cardiac risk factors. Our analysis has shown that "social factors," such as living alone, may disproportionately increase the risk of atrial fibrillation in the under-represented non-White patient group with heart failure, highlighting the need for more studies focusing on stratification of multiracial cohorts to better uncover the heterogeneity of atrial fibrillation.

Machine Learning Predicts Cardiovascular Events in Patients With Diabetes: The Silesia Diabetes-Heart Project.

We aimed to develop a machine learning (ML) model for predicting cardiovascular (CV) events in patients with diabetes (DM). This was a prospective, observational study where clinical data of patients with diabetes hospitalized in the diabetology center in Poland (years 2015-2020) were analyzed using ML. The occurrence of new CV events following discharge was collected in the follow-up time for up to 5 years and 9 months. An end-to-end ML technique which exploits the neighborhood component analysis for elaborating discriminative predictors, followed by a hybrid sampling/boosting classification algorithm, multiple logistic regression (MLR), or unsupervised hierarchical clustering was proposed. In 1735 patients with diabetes (53% female), there were 150 (8.65%) ones with a new CV event in the follow-up. Twelve most discriminative patients' parameters included coronary artery disease, heart failure, peripheral artery disease, stroke, diabetic foot disease, chronic kidney disease, eosinophil count, serum potassium level, and being treated with clopidogrel, heparin, proton pump inhibitor, and loop diuretic. Utilizing those variables resulted in the area under the receiver operating characteristic curve (AUC) ranging from 0.62 (95% Confidence Interval [CI] 0.56-0.68, P < 0.01) to 0.72 (95% CI 0.66-0.77, P < 0.01) across 5 nonoverlapping test folds, whereas MLR correctly determined 111/150 (74.00%) high-risk patients, and 989/1585 (62.40%) low-risk patients, resulting in 1100/1735 (63.40%) correctly classified patients (AUC: 0.72, 95% CI 0.66-0.77). ML algorithms can identify patients with diabetes at a high risk of new CV events based on a small number of interpretable and easy-to-obtain patients' parameters.

Improving Competency in Safe Patient Handling Through Online Learning Beyond the Classroom: A Longitudinal Study.

IMPORTANCE: Safe patient handling is intrinsic in health care provision, yet education in the skills required for safe patient handling is inconsistently delivered, with limited evidence that traditional face-to-face training reduces risk. OBJECTIVE: To assess the long-term effectiveness of replacing annual practical handling updates with an online training system, combined with competency assessment of skill and safety. DESIGN: Quasi-experimental longitudinal 3-yr study to track practical people handling skill development among undergraduate occupational therapy students. All participants had access to a multimedia online training system (that replaced tutor-led practical training), used in combination with annual competency evaluations to measure skills and safety in four people handling tasks. SETTING: All competency assessment took place on site in the School of Health and Society, University of Salford (Salford, United Kingdom). PARTICIPANTS: Undergraduate BSc(Hons) occupational therapy students (N = 243). Outcomes and Measures: Participants attended individual 45-min competency evaluations at three data collection points: beginning of Years 2 and 3 and end of Year 3. Data were collected by trained assessors using a competency assessment tool. RESULTS: Results demonstrate significant increases in skill level for sit-to-stand and repositioning in the chair (p < .05) and for hoisting and slide sheet maneuvers (p < .0001). Participants achieved 100% safety scores for repositioning in the chair and hoisting. CONCLUSIONS AND RELEVANCE: Students using the online system performed significantly better than students receiving traditional annual practical updates, providing an evidence base to reduce tutor-led training hours while increasing skill and safety levels using a combination of the online system and competency assessment. What This Article Adds: This approach was found to reinforce safe handling techniques and increase independence, competency, and safety of service users and caregivers working in health and social care environments while reducing time spent delivering annual people handling updates. The findings support replacement of face-to-face training updates, particularly in the current climate of social distancing.

Using probe electrospray ionization mass spectrometry and machine learning for detecting pancreatic cancer with high performance.

A rapid blood-based diagnostic modality to detect pancreatic ductal adenocarcinoma (PDAC) with high accuracy is an unmet medical need. The study aimed to validate a unique diagnosis system using Probe Electrospray Ionization Mass Spectrometry (PESI-MS) and Machine Learning to the diagnosis of PDAC. Peripheral blood samples were collected from a total of 322 consecutive PDAC patients and 265 controls with a family history of PDAC. Five µl of serum samples were analyzed using PESI-MS system. The mass spectra from each specimen were then fed into machine learning algorithms to discriminate between control and cancer cases. A total of 587 serum samples were analyzed. The sensitivity of the machine learning algorithm using PESI-MS profiles to identify PDAC is 90.8% with specificity of 91.7% (95% CI 83.9%-97.4% and 82.8%-97.7% respectively). Combined PESI-MS profiles with age and CA19-9 as predictors, the accuracy for stage 1 or 2 of PDAC is 92.9% and for stage 3 or 4 is 93% (95% CI 86.3-98.2; 87.9-97.4 respectively). The accuracy and simplicity of the PESI-MS profiles combined with machine learning provide an opportunity to detect PDAC at an early stage and must be applicable to the examination of at-risk populations.

pH plays a role in the mode of action of trimethoprim on Escherichia coli.

Metabolomics-based approaches were applied to understand interactions of trimethoprim with Escherichia coli K-12 at sub-minimum inhibitory concentrations (MIC≈0.2, 0.03 and 0.003 mg L-1). Trimethoprim inhibits dihydrofolate reductase and thereby is an indirect inhibitor of nucleic acid synthesis. Due to the basicity of trimethoprim, two pH levels (5 and 7) were selected which mimicked healthy urine pH. This also allowed investigation of the effect on bacterial metabolism when trimethoprim exists in different ionization states. UHPLC-MS was employed to detect trimethoprim molecules inside the bacterial cell and this showed that at pH 7 more of the drug was recovered compared to pH 5; this correlated with classical growth curve measurements. FT-IR spectroscopy was used to establish recovery of reproducible phenotypes under all 8 conditions (3 drug levels and control in 2 pH levels) and GC-MS was used to generate global metabolic profiles. In addition to finding direct mode-of-action effects where nucleotides were decreased at pH 7 with increasing trimethoprim levels, off-target pH-related effects were observed for many amino acids. Additionally, stress-related effects were observed where the osmoprotectant trehalose was higher at increased antibiotic levels at pH 7. This correlated with glucose and fructose consumption and increase in pyruvate-related products as well as lactate and alanine. Alanine is a known regulator of sugar metabolism and this increase may be to enhance sugar consumption and thus trehalose production. These results provide a wider view of the action of trimethoprim. Metabolomics indicated alternative metabolism areas to be investigated to further understand the off-target effects of trimethoprim.

Glycemia but not the Metabolic Syndrome is Associated with Cognitive Decline: Findings from the European Male Ageing Study.

OBJECTIVE: Previous research has indicated that components of the metabolic syndrome (MetS), such as hyperglycemia and hypertension, are negatively associated with cognition. However, evidence that MetS itself is related to cognitive performance has been inconsistent. This longitudinal study investigates whether MetS or its components affect cognitive decline in aging men and whether any interaction with inflammation exists. METHODS: Over a mean of 4.4 years (SD ± 0.3), men aged 40-79 years from the multicenter European Male Ageing Study were recruited. Cognitive functioning was assessed using the Rey-Osterrieth Complex Figure (ROCF), the Camden Topographical Recognition Memory (CTRM) task, and the Digit Symbol Substitution Test (DSST). High-sensitivity C-reactive protein (hs-CRP) levels were measured using a chemiluminescent immunometric assay. RESULTS: Overall, 1,913 participants contributed data to the ROCF analyses and 1,965 subjects contributed to the CTRM and DSST analyses. In multiple regression models the presence of baseline MetS was not associated with cognitive decline over time (p > 0.05). However, logistic ordinal regressions indicated that high glucose levels were related to a greater risk of decline on the ROCF Copy (ß = -0.42, p < 0.05) and the DSST (ß = -0.39, p < 0.001). There was neither a main effect of hs-CRP levels nor an interaction effect of hs-CRP and MetS at baseline on cognitive decline. CONCLUSION: No evidence was found for a relationship between MetS or inflammation and cognitive decline in this sample of aging men. However, glycemia was negatively associated with visuoconstructional abilities and processing speed.

Detection of glycosylation and iron-binding protein modifications using Raman spectroscopy.

In this study we demonstrate the use of Raman spectroscopy to determine protein modifications as a result of glycosylation and iron binding. Most proteins undergo some modifications after translation which can directly affect protein function. Identifying these modifications is particularly important in the production of biotherapeutic agents as they can affect stability, immunogenicity and pharmacokinetics. However, post-translational modifications can often be difficult to detect with regard to the subtle structural changes they induce in proteins. From their Raman spectra apo-and holo-forms of iron-binding proteins, transferrin and ferritin, could be readily distinguished and variations in spectral features as a result of structural changes could also be determined. In particular, differences in solvent exposure of aromatic amino acids residues could be identified between the open and closed forms of the iron-binding proteins. Protein modifications as a result of glycosylation can be even more difficult to identify. Through the application of the chemometric techniques of principal component analysis and partial least squares regression variations in Raman spectral features as a result of glycosylation induced structural modifications could be identified. These were then used to distinguish between glycosylated and non-glycosylated transferrin and to measure the relative concentrations of the glycoprotein within a mixture of the native non-glycosylated protein.

Evaluation of cognitive subdomains, 25-hydroxyvitamin D, and 1,25-dihydroxyvitamin D in the European Male Ageing Study.

PURPOSE: Although lower levels of vitamin D have been related to poor cognitive functioning and dementia in older adults, evidence from longitudinal investigations is inconsistent. The objective of this study was to determine whether 25-hydroxyvitamin D [25(OH)D] and 1,25-dihydroxyvitamin D [1,25(OH)2D] levels are associated with specified measures of cognitive decline in ageing men. METHODS: The European Male Ageing Study (EMAS) followed 3369 men aged 40-79 over 4.4 years. 25(OH)D levels at baseline were measured by radioimmunoassay, and 1,25(OH)2D levels were obtained with liquid chromatography-tandem mass spectrometry. Visuoconstructional abilities, visual memory, and processing speed at baseline and follow-up were assessed using the Rey-Osterrieth Complex Figure Test (ROCF), Camden Topographical Recognition Memory (CTRM), and the Digit Symbol Substitution Test (DSST). RESULTS: Following attritions, a total of 2430 men with a mean (SD) age of 59.0 (10.6) were included in the analyses. At baseline, the mean 25(OH)D concentration was 64.6 (31.5) nmol/l, and mean 1,25(OH)2D level was 59.6 (16.6) pmol/l. In age-adjusted linear regression models, high 25(OH)D concentrations were associated with a smaller decline in the DSST (ß = 0.007, p = 0.020). Men with low 25(OH)D levels (<50 nmol/l) showed a greater decline in the CTRM compared to men with higher (≥75 nmol/l) levels (ß = -0.41, p = 0.035). However, these associations disappeared after adjusting for confounders such as depressive symptoms, BMI, and comorbidities. There was no indication of a relationship between 1,25(OH)2D and decline in cognitive subdomains. CONCLUSION: We found no evidence for an independent association between 25(OH)D or 1,25(OH)2D levels and visuoconstructional abilities, visual memory, or processing speed over on average 4.4 years in this sample of middle-aged and elderly European men.

Changes in prevalence of obesity and high waist circumference over four years across European regions: the European male ageing study (EMAS).

Diversity in lifestyles and socioeconomic status among European populations, and recent socio-political and economic changes in transitional countries, may affect changes in adiposity. We aimed to determine whether change in the prevalence of obesity varies between the socio-politically transitional North-East European (Lódz, Poland; Szeged, Hungary; Tartu, Estonia), and the non-transitional Mediterranean (Santiago de Compostela, Spain; Florence, Italy) and North-West European (Leuven, Belgium; Malmö, Sweden; Manchester, UK) cities. This prospective observational cohort survey was performed between 2003 and 2005 at baseline and followed up between 2008 and 2010 of 3369 community-dwelling men aged 40-79 years. Main outcome measures in the present paper included waist circumference, body mass index and mid-upper arm muscle area. Baseline prevalence of waist circumference ≥ 102 cm and body mass index ≥ 30 kg/m2, respectively, were 39.0, 29.5 % in North-East European cities, 32.4, 21.9 % in Mediterranean cities, and 30.0, 20.1 % in North-West European cities. After median 4.3 years, men living in cities from transitional countries had mean gains in waist circumference (1.1 cm) and body mass index (0.2 kg/m2), which were greater than men in cities from non-transitional countries (P = 0.005). North-East European cities had greater gains in waist circumference (1.5 cm) than in Mediterranean cities (P < 0.001). Over 4.3 years, the prevalence of waist circumference ≥ 102 cm had increased by 13.1 % in North-East European cities, 5.8 % in the Mediterranean cities, 10.0 % in North-West European cities. Odds ratios (95 % confidence intervals), adjusted for lifestyle factors, for developing waist circumference ≥ 102 cm, compared with men from Mediterranean cities, were 2.3 (1.5-3.5) in North-East European cities and 1.6 (1.1-2.4) in North-West European cities, and 1.6 (1.2-2.1) in men living in cities from transitional, compared with cities from non-transitional countries. These regional differences in increased prevalence of waist circumference ≥ 102 cm were more pronounced in men aged 60-79 years than in those aged 40-59 years. Overall there was an increase in the prevalence of obesity (body mass index ≥ 30 kg/m2) over 4.3 years (between 5.3 and 6.1 %) with no significant regional differences at any age. Mid-upper arm muscle area declined during follow-up with the greatest decline among men from North-East European cities. In conclusion, increasing waist circumference is dissociated from change in body mass index and most rapid among men living in cities from transitional North-East European countries, presumably driven by economic and socio-political changes. Information on women would also be of value and it would be of interest to relate the changes in adiposity to dietary and other behavioural habits.

Fractional Factorial Design of MALDI-TOF-MS Sample Preparations for the Optimized Detection of Phospholipids and Acylglycerols.

Matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF-MS) has successfully been used for the analysis of high molecular weight compounds, such as proteins and nucleic acids. By contrast, analysis of low molecular weight compounds with this technique has been less successful due to interference from matrix peaks which have a similar mass to the target analyte(s). Recently, a variety of modified matrices and matrix additives have been used to overcome these limitations. An increased interest in lipid analysis arose from the feasibility of correlating these components with many diseases, e.g. atherosclerosis and metabolic dysfunctions. Lipids have a wide range of chemical properties making their analysis difficult with traditional methods. MALDI-TOF-MS shows excellent potential for sensitive and rapid analysis of lipids, and therefore this study focuses on computational-analytical optimization of the analysis of five lipids (4 phospholipids and 1 acylglycerol) in complex mixtures using MALDI-TOF-MS with fractional factorial design (FFD) and Pareto optimality. Five different experimental factors were investigated using FFD which reduced the number of experiments performed by identifying 720 key experiments from a total of 8064 possible analyses. Factors investigated included the following: matrices, matrix preparations, matrix additives, additive concentrations, and deposition methods. This led to a significant reduction in time and cost of sample analysis with near optimal conditions. We discovered that the key factors used to produce high quality spectra were the matrix and use of appropriate matrix additives.

Data standards can boost metabolomics research, and if there is a will, there is a way.

Thousands of articles using metabolomics approaches are published every year. With the increasing amounts of data being produced, mere description of investigations as text in manuscripts is not sufficient to enable re-use anymore: the underlying data needs to be published together with the findings in the literature to maximise the benefit from public and private expenditure and to take advantage of an enormous opportunity to improve scientific reproducibility in metabolomics and cognate disciplines. Reporting recommendations in metabolomics started to emerge about a decade ago and were mostly concerned with inventories of the information that had to be reported in the literature for consistency. In recent years, metabolomics data standards have developed extensively, to include the primary research data, derived results and the experimental description and importantly the metadata in a machine-readable way. This includes vendor independent data standards such as mzML for mass spectrometry and nmrML for NMR raw data that have both enabled the development of advanced data processing algorithms by the scientific community. Standards such as ISA-Tab cover essential metadata, including the experimental design, the applied protocols, association between samples, data files and the experimental factors for further statistical analysis. Altogether, they pave the way for both reproducible research and data reuse, including meta-analyses. Further incentives to prepare standards compliant data sets include new opportunities to publish data sets, but also require a little "arm twisting" in the author guidelines of scientific journals to submit the data sets to public repositories such as the NIH Metabolomics Workbench or MetaboLights at EMBL-EBI. In the present article, we look at standards for data sharing, investigate their impact in metabolomics and give suggestions to improve their adoption.

COordination of Standards in MetabOlomicS (COSMOS): facilitating integrated metabolomics data access.

Metabolomics has become a crucial phenotyping technique in a range of research fields including medicine, the life sciences, biotechnology and the environmental sciences. This necessitates the transfer of experimental information between research groups, as well as potentially to publishers and funders. After the initial efforts of the metabolomics standards initiative, minimum reporting standards were proposed which included the concepts for metabolomics databases. Built by the community, standards and infrastructure for metabolomics are still needed to allow storage, exchange, comparison and re-utilization of metabolomics data. The Framework Programme 7 EU Initiative 'coordination of standards in metabolomics' (COSMOS) is developing a robust data infrastructure and exchange standards for metabolomics data and metadata. This is to support workflows for a broad range of metabolomics applications within the European metabolomics community and the wider metabolomics and biomedical communities' participation. Here we announce our concepts and efforts asking for re-engagement of the metabolomics community, academics and industry, journal publishers, software and hardware vendors, as well as those interested in standardisation worldwide (addressing missing metabolomics ontologies, complex-metadata capturing and XML based open source data exchange format), to join and work towards updating and implementing metabolomics standards.

Phenotypic Characterisation of Shewanella oneidensis MR-1 Exposed to X-Radiation.

Biogeochemical processes mediated by Fe(III)-reducing bacteria such as Shewanella oneidensis have the potential to influence the post-closure evolution of a geological disposal facility for radioactive wastes and to affect the solubility of some radionuclides. Furthermore, their potential to reduce both Fe(III) and radionuclides can be harnessed for the bioremediation of radionuclide-contaminated land. As some such sites are likely to have significant radiation fluxes, there is a need to characterise the impact of radiation stress on such microorganisms. There have, however, been few global cell analyses on the impact of ionizing radiation on subsurface bacteria, so here we address the metabolic response of S. oneidensis MR-1 to acute doses of X-radiation. UV/Vis spectroscopy and CFU counts showed that although X-radiation decreased initial viability and extended the lag phase of batch cultures, final biomass yields remained unchanged. FT-IR spectroscopy of whole cells indicated an increase in lipid associated vibrations and decreases in vibrations tentatively assigned to nucleic acids, phosphate, saccharides and amines. MALDI-TOF-MS detected an increase in total protein expression in cultures exposed to 12 Gy. At 95 Gy, a decrease in total protein levels was generally observed, although an increase in a putative cold shock protein was observed, which may be related to the radiation stress response of this organism. Multivariate statistical analyses applied to these FT-IR and MALDI-TOF-MS spectral data suggested that an irradiated phenotype developed throughout subsequent generations. This study suggests that significant alteration to the metabolism of S. oneidensis MR-1 is incurred as a result of X-irradiation and that dose dependent changes to specific biomolecules characterise this response. Irradiated S. oneidensis also displayed enhanced levels of poorly crystalline Fe(III) oxide reduction, though the mechanism underpinning this phenomenon is unclear.

A tutorial review: Metabolomics and partial least squares-discriminant analysis--a marriage of convenience or a shotgun wedding.

The predominance of partial least squares-discriminant analysis (PLS-DA) used to analyze metabolomics datasets (indeed, it is the most well-known tool to perform classification and regression in metabolomics), can be said to have led to the point that not all researchers are fully aware of alternative multivariate classification algorithms. This may in part be due to the widespread availability of PLS-DA in most of the well-known statistical software packages, where its implementation is very easy if the default settings are used. In addition, one of the perceived advantages of PLS-DA is that it has the ability to analyze highly collinear and noisy data. Furthermore, the calibration model is known to provide a variety of useful statistics, such as prediction accuracy as well as scores and loadings plots. However, this method may provide misleading results, largely due to a lack of suitable statistical validation, when used by non-experts who are not aware of its potential limitations when used in conjunction with metabolomics. This tutorial review aims to provide an introductory overview to several straightforward statistical methods such as principal component-discriminant function analysis (PC-DFA), support vector machines (SVM) and random forests (RF), which could very easily be used either to augment PLS or as alternative supervised learning methods to PLS-DA. These methods can be said to be particularly appropriate for the analysis of large, highly-complex data sets which are common output(s) in metabolomics studies where the numbers of variables often far exceed the number of samples. In addition, these alternative techniques may be useful tools for generating parsimonious models through feature selection and data reduction, as well as providing more propitious results. We sincerely hope that the general reader is left with little doubt that there are several promising and readily available alternatives to PLS-DA, to analyze large and highly complex data sets.

Oxidized phosphatidylcholines suggest oxidative stress in patients with medium-chain acyl-CoA dehydrogenase deficiency.

Inborn errors of metabolism encompass a large group of diseases caused by enzyme deficiencies and are therefore amenable to metabolomics investigations. Medium chain acyl-CoA dehydrogenase deficiency (MCADD) is a defect in ß-oxidation of fatty acids, and is one of the most well understood disorders. We report here the use of liquid chromatography-mass spectrometry (LC-MS) based untargeted metabolomics and targeted flow injection analysis-tandem mass spectrometry (FIA-TMS) that lead to discovery of novel compounds of oxidative stress. Dry blood spots of controls (n=25) and patient samples (n=25) were extracted by methanol/water (1/1, v/v) and these supernatants were analyzed by LC-MS method with detection by an Orbitrap Elite MS. Data were processed by XCMS and CAMERA followed by dimension reduction methods. Patients were clearly distinguished from controls in PCA. S-plot derived from OPLS-DA indicated that medium-chain acylcarnitines (octanoyl, decenoyl and decanoyl carnitines) as well as three phosphatidylcholines (PC(16:0,9:0(COOH))), PC(18:0,5:0(COOH)) and PC(16:0,8:0(COOH)) were important metabolites for differentiation between patients and healthy controls. In order to biologically validate these discriminatory molecules as indicators for oxidative stress, a second cohort of individuals were analyzed, including MCADD (n=25) and control (n=250) samples. These were measured by a modified newborn screening method using FIA-TMS (API 4000) in MRM mode. Calculated p-values for PC(16:0,9:0(COOH)), PC(18:0,5:0(COOH)) and PC(16:0,8:0(COOH)) were 1.927×10(-14), 2.391×10(-15) and 3.354×10(-15) respectively. These elevated oxidized phospholipids indeed show an increased presence of oxidative stress in MCADD patients as one of the pathophysiological mechanisms of the disease.

Metabolic Profiling of Geobacter sulfurreducens during Industrial Bioprocess Scale-Up.

During the industrial scale-up of bioprocesses it is important to establish that the biological system has not changed significantly when moving from small laboratory-scale shake flasks or culturing bottles to an industrially relevant production level. Therefore, during upscaling of biomass production for a range of metal transformations, including the production of biogenic magnetite nanoparticles by Geobacter sulfurreducens, from 100-ml bench-scale to 5-liter fermentors, we applied Fourier transform infrared (FTIR) spectroscopy as a metabolic fingerprinting approach followed by the analysis of bacterial cell extracts by gas chromatography-mass spectrometry (GC-MS) for metabolic profiling. FTIR results clearly differentiated between the phenotypic changes associated with different growth phases as well as the two culturing conditions. Furthermore, the clustering patterns displayed by multivariate analysis were in agreement with the turbidimetric measurements, which displayed an extended lag phase for cells grown in a 5-liter bioreactor (24 h) compared to those grown in 100-ml serum bottles (6 h). GC-MS analysis of the cell extracts demonstrated an overall accumulation of fumarate during the lag phase under both culturing conditions, coinciding with the detected concentrations of oxaloacetate, pyruvate, nicotinamide, and glycerol-3-phosphate being at their lowest levels compared to other growth phases. These metabolites were overlaid onto a metabolic network of G. sulfurreducens, and taking into account the levels of these metabolites throughout the fermentation process, the limited availability of oxaloacetate and nicotinamide would seem to be the main metabolic bottleneck resulting from this scale-up process. Additional metabolite-feeding experiments were carried out to validate the above hypothesis. Nicotinamide supplementation (1 mM) did not display any significant effects on the lag phase of G. sulfurreducens cells grown in the 100-ml serum bottles. However, it significantly improved the growth behavior of cells grown in the 5-liter bioreactor by reducing the lag phase from 24 h to 6 h, while providing higher yield than in the 100-ml serum bottles.

Profiling of spatial metabolite distributions in wheat leaves under normal and nitrate limiting conditions.

The control and interaction between nitrogen and carbon assimilatory pathways is essential in both photosynthetic and non-photosynthetic tissue in order to support metabolic processes without compromising growth. Physiological differences between the basal and mature region of wheat (Triticum aestivum) primary leaves confirmed that there was a change from heterotrophic to autotrophic metabolism. Fourier Transform Infrared (FT-IR) spectroscopy confirmed the suitability and phenotypic reproducibility of the leaf growth conditions. Principal Component-Discriminant Function Analysis (PC-DFA) revealed distinct clustering between base, and tip sections of the developing wheat leaf, and from plants grown in the presence or absence of nitrate. Gas Chromatography-Time of Flight/Mass Spectrometry (GC-TOF/MS) combined with multivariate and univariate analyses, and Bayesian network (BN) analysis, distinguished different tissues and confirmed the physiological switch from high rates of respiration to photosynthesis along the leaf. The operation of nitrogen metabolism impacted on the levels and distribution of amino acids, organic acids and carbohydrates within the wheat leaf. In plants grown in the presence of nitrate there was reduced levels of a number of sugar metabolites in the leaf base and an increase in maltose levels, possibly reflecting an increase in starch turnover. The value of using this combined metabolomics analysis for further functional investigations in the future are discussed.

A comparison of different chemometrics approaches for the robust classification of electronic nose data.

Accurate detection of certain chemical vapours is important, as these may be diagnostic for the presence of weapons, drugs of misuse or disease. In order to achieve this, chemical sensors could be deployed remotely. However, the readout from such sensors is a multivariate pattern, and this needs to be interpreted robustly using powerful supervised learning methods. Therefore, in this study, we compared the classification accuracy of four pattern recognition algorithms which include linear discriminant analysis (LDA), partial least squares-discriminant analysis (PLS-DA), random forests (RF) and support vector machines (SVM) which employed four different kernels. For this purpose, we have used electronic nose (e-nose) sensor data (Wedge et al., Sensors Actuators B Chem 143:365-372, 2009). In order to allow direct comparison between our four different algorithms, we employed two model validation procedures based on either 10-fold cross-validation or bootstrapping. The results show that LDA (91.56% accuracy) and SVM with a polynomial kernel (91.66% accuracy) were very effective at analysing these e-nose data. These two models gave superior prediction accuracy, sensitivity and specificity in comparison to the other techniques employed. With respect to the e-nose sensor data studied here, our findings recommend that SVM with a polynomial kernel should be favoured as a classification method over the other statistical models that we assessed. SVM with non-linear kernels have the advantage that they can be used for classifying non-linear as well as linear mapping from analytical data space to multi-group classifications and would thus be a suitable algorithm for the analysis of most e-nose sensor data.

Influence of missing values substitutes on multivariate analysis of metabolomics data.

Missing values are known to be problematic for the analysis of gas chromatography-mass spectrometry (GC-MS) metabolomics data. Typically these values cover about 10%-20% of all data and can originate from various backgrounds, including analytical, computational, as well as biological. Currently, the most well known substitute for missing values is a mean imputation. In fact, some researchers consider this aspect of data analysis in their metabolomics pipeline as so routine that they do not even mention using this replacement approach. However, this may have a significant influence on the data analysis output(s) and might be highly sensitive to the distribution of samples between different classes. Therefore, in this study we have analysed different substitutes of missing values namely: zero, mean, median, k-nearest neighbours (kNN) and random forest (RF) imputation, in terms of their influence on unsupervised and supervised learning and, thus, their impact on the final output(s) in terms of biological interpretation. These comparisons have been demonstrated both visually and computationally (classification rate) to support our findings. The results show that the selection of the replacement methods to impute missing values may have a considerable effect on the classification accuracy, if performed incorrectly this may negatively influence the biomarkers selected for an early disease diagnosis or identification of cancer related metabolites. In the case of GC-MS metabolomics data studied here our findings recommend that RF should be favored as an imputation of missing value over the other tested methods. This approach displayed excellent results in terms of classification rate for both supervised methods namely: principal components-linear discriminant analysis (PC-LDA) (98.02%) and partial least squares-discriminant analysis (PLS-DA) (97.96%) outperforming other imputation methods.

A comparative investigation of modern feature selection and classification approaches for the analysis of mass spectrometry data.

Many analytical approaches such as mass spectrometry generate large amounts of data (input variables) per sample analysed, and not all of these variables are important or related to the target output of interest. The selection of a smaller number of variables prior to sample classification is a widespread task in many research studies, where attempts are made to seek the lowest possible set of variables that are still able to achieve a high level of prediction accuracy; in other words, there is a need to generate the most parsimonious solution when the number of input variables is huge but the number of samples/objects are smaller. Here, we compare several different variable selection approaches in order to ascertain which of these are ideally suited to achieve this goal. All variable selection approaches were applied to the analysis of a common set of metabolomics data generated by Curie-point pyrolysis mass spectrometry (Py-MS), where the goal of the study was to classify the Gram-positive bacteria Bacillus. These approaches include stepwise forward variable selection, used for linear discriminant analysis (LDA); variable importance for projection (VIP) coefficient, employed in partial least squares-discriminant analysis (PLS-DA); support vector machines-recursive feature elimination (SVM-RFE); as well as the mean decrease in accuracy and mean decrease in Gini, provided by random forests (RF). Finally, a double cross-validation procedure was applied to minimize the consequence of overfitting. The results revealed that RF with its variable selection techniques and SVM combined with SVM-RFE as a variable selection method, displayed the best results in comparison to other approaches.