Multiple Kernel Learning of Environmental Data. Case study: Analysis and Mapping of Wind Fields

The paper presents the Multiple Kernel Learning (MKL) approach as a modelling and data exploratory tool and applies it to the problem of wind speed mapping. Support Vector Regression (SVR) is used to predict spatial variations of the mean wind speed from terrain features (slopes, terrain curvature, directional derivatives) generated at different spatial scales. Multiple Kernel Learning is applied to learn kernels for individual features and thematic feature subsets, both in the context of feature selection and optimal parameters determination. An empirical study on real-life data confirms the usefulness of MKL as a tool that enhances the interpretability of data-driven models.

New microfluidic-based sampling procedure for overcoming the hematocrit problem associated with dried blood spot analysis.

Hematocrit (Hct) is one of the most critical issues associated with the bioanalytical methods used for dried blood spot (DBS) sample analysis. Because Hct determines the viscosity of blood, it may affect the spreading of blood onto the filter paper. Hence, accurate quantitative data can only be obtained if the size of the paper filter extracted contains a fixed blood volume. We describe for the first time a microfluidic-based sampling procedure to enable accurate blood volume collection on commercially available DBS cards. The system allows the collection of a controlled volume of blood (e.g., 5 or 10 μL) within several seconds. Reproducibility of the sampling volume was examined in vivo on capillary blood by quantifying caffeine and paraxanthine on 5 different extracted DBS spots at two different time points and in vitro with a test compound, Mavoglurant, on 10 different spots at two Hct levels. Entire spots were extracted. In addition, the accuracy and precision (n = 3) data for the Mavoglurant quantitation in blood with Hct levels between 26% and 62% were evaluated. The interspot precision data were below 9.0%, which was equivalent to that of a manually spotted volume with a pipet. No Hct effect was observed in the quantitative results obtained for Hct levels from 26% to 62%. These data indicate that our microfluidic-based sampling procedure is accurate and precise and that the analysis of Mavoglurant is not affected by the Hct values. This provides a simple procedure for DBS sampling with a fixed volume of capillary blood, which could eliminate the recurrent Hct issue linked to DBS sample analysis.

Individual prediction of cognitive decline in mild cognitive impairment using support vector machine-based analysis of diffusion tensor imaging data.

Although cross-sectional diffusion tensor imaging (DTI) studies revealed significant white matter changes in mild cognitive impairment (MCI), the utility of this technique in predicting further cognitive decline is debated. Thirty-five healthy controls (HC) and 67 MCI subjects with DTI baseline data were neuropsychologically assessed at one year. Among them, there were 40 stable (sMCI; 9 single domain amnestic, 7 single domain frontal, 24 multiple domain) and 27 were progressive (pMCI; 7 single domain amnestic, 4 single domain frontal, 16 multiple domain). Fractional anisotropy (FA) and longitudinal, radial, and mean diffusivity were measured using Tract-Based Spatial Statistics. Statistics included group comparisons and individual classification of MCI cases using support vector machines (SVM). FA was significantly higher in HC compared to MCI in a distributed network including the ventral part of the corpus callosum, right temporal and frontal pathways. There were no significant group-level differences between sMCI versus pMCI or between MCI subtypes after correction for multiple comparisons. However, SVM analysis allowed for an individual classification with accuracies up to 91.4% (HC versus MCI) and 98.4% (sMCI versus pMCI). When considering the MCI subgroups separately, the minimum SVM classification accuracy for stable versus progressive cognitive decline was 97.5% in the multiple domain MCI group. SVM analysis of DTI data provided highly accurate individual classification of stable versus progressive MCI regardless of MCI subtype, indicating that this method may become an easily applicable tool for early individual detection of MCI subjects evolving to dementia.

The Potential of Friction as a Tool for Winter Maintenance, TR-400, 1998

It is intuitively obvious that snow or ice on a road surface will make that surface more slippery and thus more hazardous. However, quantifying this slipperiness by measuring the friction between the road surface and a vehicle is rather difficult. If such friction readings could be easily made, they might provide a means to control winter maintenance activities more efficiently than at present. This study is a preliminary examination of the possibility of using friction as an operational tool in winter maintenance. In particular, the relationship of friction to traffic volume and speed, and accident rates is examined, and the current lack of knowledge in this area is outlined. The state of the art of friction measuring techniques is reviewed. A series of experiments whereby greater knowledge of how friction deteriorates during a storm and is restored by treatment is proposed. The relationship between plowing forces and the ice-pavement bond strength is discussed. The challenge of integrating all these potential sources of information into a useful final product is presented together with a potential approach. A preliminary cost-benefit analysis of friction measuring devices is performed and suggests that considerable savings might be realized if certain assumptions should hold true. The steps required to bring friction from its current state as a research tool to full deployment as an operational tool are presented and discussed. While much remains to be done in this regard, it is apparent that friction could be an extremely effective operational tool in winter maintenance activities of the future.

Experimental analysis and computer prediction of CTF/NFI transcription factor DNA binding sites.

Accurate prediction of transcription factor binding sites is needed to unravel the function and regulation of genes discovered in genome sequencing projects. To evaluate current computer prediction tools, we have begun a systematic study of the sequence-specific DNA-binding of a transcription factor belonging to the CTF/NFI family. Using a systematic collection of rationally designed oligonucleotides combined with an in vitro DNA binding assay, we found that the sequence specificity of this protein cannot be represented by a simple consensus sequence or weight matrix. For instance, CTF/NFI uses a flexible DNA binding mode that allows for variations of the binding site length. From the experimental data, we derived a novel prediction method using a generalised profile as a binding site predictor. Experimental evaluation of the generalised profile indicated that it accurately predicts the binding affinity of the transcription factor to natural or synthetic DNA sequences. Furthermore, the in vitro measured binding affinities of a subset of oligonucleotides were found to correlate with their transcriptional activities in transfected cells. The combined computational-experimental approach exemplified in this work thus resulted in an accurate prediction method for CTF/NFI binding sites potentially functioning as regulatory regions in vivo.

A Statistical Analysis of Iowa Premature Concrete Pavement Deterioration, Supplementary Study, HR-1063, 1997

This supplementary project has been undertaken as an effort to continue work previously completed in the Pooled Fund Study of Premature Concrete Pavement Deterioration. As such, it shares the objective of "Identifying the variables that are present in those pavements exhibiting premature deterioration," by collecting additional data and performing statistical analysis of those data. The approach and philosophy of this work are identical to that followed in the above project, and the Pooled Fund Study Final Report provides a detailed description of this process. This project has involved the collection of data for additional sites in the state of Iowa. These sites have then been added to sites collected in the original study, and statistical analysis has been performed on the entire set. It is hoped that this will have two major effects. First, using data from only one state allows for the analysis of a larger set of independent variables with a greater degree of commonality than was possible in the multi-state study, since the data are not limited by state to state differences in data collection and retention. Second, more data on additional sites will increase the degrees of freedom in the model and hopefully add confidence to the results.

ConnectomeViewer - Multi-Modal Multi-Level Network Visualization and Analysis.

Introduction: The field of Connectomic research is growing rapidly, resulting from methodological advances in structural neuroimaging on many spatial scales. Especially progress in Diffusion MRI data acquisition and processing made available macroscopic structural connectivity maps in vivo through Connectome Mapping Pipelines (Hagmann et al, 2008) into so-called Connectomes (Hagmann 2005, Sporns et al, 2005). They exhibit both spatial and topological information that constrain functional imaging studies and are relevant in their interpretation. The need for a special-purpose software tool for both clinical researchers and neuroscientists to support investigations of such connectome data has grown. Methods: We developed the ConnectomeViewer, a powerful, extensible software tool for visualization and analysis in connectomic research. It uses the novel defined container-like Connectome File Format, specifying networks (GraphML), surfaces (Gifti), volumes (Nifti), track data (TrackVis) and metadata. Usage of Python as programming language allows it to by cross-platform and have access to a multitude of scientific libraries. Results: Using a flexible plugin architecture, it is possible to enhance functionality for specific purposes easily. Following features are already implemented: * Ready usage of libraries, e.g. for complex network analysis (NetworkX) and data plotting (Matplotlib). More brain connectivity measures will be implemented in a future release (Rubinov et al, 2009). * 3D View of networks with node positioning based on corresponding ROI surface patch. Other layouts possible. * Picking functionality to select nodes, select edges, get more node information (ConnectomeWiki), toggle surface representations * Interactive thresholding and modality selection of edge properties using filters * Arbitrary metadata can be stored for networks, thereby allowing e.g. group-based analysis or meta-analysis. * Python Shell for scripting. Application data is exposed and can be modified or used for further post-processing. * Visualization pipelines using filters and modules can be composed with Mayavi (Ramachandran et al, 2008). * Interface to TrackVis to visualize track data. Selected nodes are converted to ROIs for fiber filtering The Connectome Mapping Pipeline (Hagmann et al, 2008) processed 20 healthy subjects into an average Connectome dataset. The Figures show the ConnectomeViewer user interface using this dataset. Connections are shown that occur in all 20 subjects. The dataset is freely available from the homepage (connectomeviewer.org). Conclusions: The ConnectomeViewer is a cross-platform, open-source software tool that provides extensive visualization and analysis capabilities for connectomic research. It has a modular architecture, integrates relevant datatypes and is completely scriptable. Visit www.connectomics.org to get involved as user or developer.

Raman spectroscopy as a powerful analytical tool: probing the structure of matter

Podeu consultar el llibre complet a: http://hdl.handle.net/2445/32166

Predictive analysis and mapping of indoor radon concentrations in Switzerland

It is estimated that around 230 people die each year due to radon (222Rn) exposure in Switzerland. 222Rn occurs mainly in closed environments like buildings and originates primarily from the subjacent ground. Therefore it depends strongly on geology and shows substantial regional variations. Correct identification of these regional variations would lead to substantial reduction of 222Rn exposure of the population based on appropriate construction of new and mitigation of already existing buildings. Prediction of indoor 222Rn concentrations (IRC) and identification of 222Rn prone areas is however difficult since IRC depend on a variety of different variables like building characteristics, meteorology, geology and anthropogenic factors. The present work aims at the development of predictive models and the understanding of IRC in Switzerland, taking into account a maximum of information in order to minimize the prediction uncertainty. The predictive maps will be used as a decision-support tool for 222Rn risk management. The construction of these models is based on different data-driven statistical methods, in combination with geographical information systems (GIS). In a first phase we performed univariate analysis of IRC for different variables, namely the detector type, building category, foundation, year of construction, the average outdoor temperature during measurement, altitude and lithology. All variables showed significant associations to IRC. Buildings constructed after 1900 showed significantly lower IRC compared to earlier constructions. We observed a further drop of IRC after 1970. In addition to that, we found an association of IRC with altitude. With regard to lithology, we observed the lowest IRC in sedimentary rocks (excluding carbonates) and sediments and the highest IRC in the Jura carbonates and igneous rock. The IRC data was systematically analyzed for potential bias due to spatially unbalanced sampling of measurements. In order to facilitate the modeling and the interpretation of the influence of geology on IRC, we developed an algorithm based on k-medoids clustering which permits to define coherent geological classes in terms of IRC. We performed a soil gas 222Rn concentration (SRC) measurement campaign in order to determine the predictive power of SRC with respect to IRC. We found that the use of SRC is limited for IRC prediction. The second part of the project was dedicated to predictive mapping of IRC using models which take into account the multidimensionality of the process of 222Rn entry into buildings. We used kernel regression and ensemble regression tree for this purpose. We could explain up to 33% of the variance of the log transformed IRC all over Switzerland. This is a good performance compared to former attempts of IRC modeling in Switzerland. As predictor variables we considered geographical coordinates, altitude, outdoor temperature, building type, foundation, year of construction and detector type. Ensemble regression trees like random forests allow to determine the role of each IRC predictor in a multidimensional setting. We found spatial information like geology, altitude and coordinates to have stronger influences on IRC than building related variables like foundation type, building type and year of construction. Based on kernel estimation we developed an approach to determine the local probability of IRC to exceed 300 Bq/m3. In addition to that we developed a confidence index in order to provide an estimate of uncertainty of the map. All methods allow an easy creation of tailor-made maps for different building characteristics. Our work is an essential step towards a 222Rn risk assessment which accounts at the same time for different architectural situations as well as geological and geographical conditions. For the communication of 222Rn hazard to the population we recommend to make use of the probability map based on kernel estimation. The communication of 222Rn hazard could for example be implemented via a web interface where the users specify the characteristics and coordinates of their home in order to obtain the probability to be above a given IRC with a corresponding index of confidence. Taking into account the health effects of 222Rn, our results have the potential to substantially improve the estimation of the effective dose from 222Rn delivered to the Swiss population.

Flow cytometry in the analysis of cellular populations

Flow cytometry has become a valuable tool in cell biology. By analyzing large number of cells individually using light-scatter and fluorescence measurements, this technique reveals both cellular characteristics and the levels of cellular components. Flow cytometry has been developed to rapidly enumerate cells and to distinguish among different cell stages and structures using multiple staining. In addition to high-speed multiparametric data acquisition, analysis and cell sorting, which allow other characteristics of individual cells to be studied, have increased the interest of researchers in this technique. This chapter gives an overview of the principles of flow cytometry and examples of the application ofthe technique.

Optimization of HS-SPME/GC-MS analysis and its use in the profiling of illicit ecstasy tablets (Part 1)

A headspace solid-phase microextraction procedure (HS-SPME) was developed for the profiling of traces present in 3,4-methylenedioxymethylampethamine (MDMA). Traces were first extracted using HS-SPME and then analyzed by gas chromatography-mass spectroscopy (GC-MS). The HS-SPME conditions were optimized using varying conditions. Optimal results were obtained when 40 mg of crushed MDMA sample was heated at 80 °C for 15 min, followed by extraction at 80 °C for 15 min with a polydimethylsiloxane/divinylbenzene coated fibre. A total of 31 compounds were identified as traces related to MDMA synthesis, namely precursors, intermediates or by-products. In addition some fatty acids used as tabletting materials and caffeine used as adulterant, were also detected. The use of a restricted set of 10 target compounds was also proposed for developing a screening tool for clustering samples having close profile. 114 seizures were analyzed using an SPME auto-sampler (MultiPurpose Samples MPS2), purchased from Gerstel GMBH & Co. (Germany), and coupled to GC-MS. The data was handled using various pre-treatment methods, followed by the study of similarities between sample pairs based on the Pearson correlation. The results show that HS-SPME, coupled with the suitable statistical method is a powerful tool for distinguishing specimens coming from the same seizure and specimens coming from different seizures. This information can be used by law enforcement personnel to visualize the ecstasy distribution network as well as the clandestine tablet manufacturing.

Fast analysis of doping agents in urine by ultra-high-pressure liquid chromatography-quadrupole time-of-flight mass spectrometry I. Screening analysis.

The general strategy to perform anti-doping analyses of urine samples starts with the screening for a wide range of compounds. This step should be fast, generic and able to detect any sample that may contain a prohibited substance while avoiding false negatives and reducing false positive results. The experiments presented in this work were based on ultra-high-pressure liquid chromatography coupled to hybrid quadrupole time-of-flight mass spectrometry. Thanks to the high sensitivity of the method, urine samples could be diluted 2-fold prior to injection. One hundred and three forbidden substances from various classes (such as stimulants, diuretics, narcotics, anti-estrogens) were analysed on a C(18) reversed-phase column in two gradients of 9min (including two 3min equilibration periods) for positive and negative electrospray ionisation and detected in the MS full scan mode. The automatic identification of analytes was based on retention time and mass accuracy, with an automated tool for peak picking. The method was validated according to the International Standard for Laboratories described in the World Anti-Doping Code and was selective enough to comply with the World Anti-Doping Agency recommendations. In addition, the matrix effect on MS response was measured on all investigated analytes spiked in urine samples. The limits of detection ranged from 1 to 500ng/mL, allowing the identification of all tested compounds in urine. When a sample was reported positive during the screening, a fast additional pre-confirmatory step was performed to reduce the number of confirmatory analyses.

Analysis of variance of primary data on plant growth analysis

Plant growth analysis presents difficulties related to statistical comparison of growth rates, and the analysis of variance of primary data could guide the interpretation of results. The objective of this work was to evaluate the analysis of variance of data from distinct harvests of an experiment, focusing especially on the homogeneity of variances and the choice of an adequate ANOVA model. Data from five experiments covering different crops and growth conditions were used. From the total number of variables, 19% were originally homoscedastic, 60% became homoscedastic after logarithmic transformation, and 21% remained heteroscedastic after transformation. Data transformation did not affect the F test in one experiment, whereas in the other experiments transformation modified the F test usually reducing the number of significant effects. Even when transformation has not altered the F test, mean comparisons led to divergent interpretations. The mixed ANOVA model, considering harvest as a random effect, reduced the number of significant effects of every factor which had the F test modified by this model. Examples illustrated that analysis of variance of primary variables provides a tool for identifying significant differences in growth rates. The analysis of variance imposes restrictions to experimental design thereby eliminating some advantages of the functional growth analysis.

Surface-enhanced laser desorption/ionisation time-of-flight mass spectrometry: A tool to predict pork quality

Expression of water soluble proteins of fresh pork Longissimus thoracis from 4 pure breed pigs (Duroc, Large White, Landrace, and Piétrain) was studied to identify candidate protein markers for meat quality. Surface-enhanced laser desorption/ionisation time-of-flight mass spectrometry (SELDI-TOF-MS) was used to obtain the soluble protein profiles of Longissimus thoracis muscles. The pure breeds showed differences among the studied meat quality traits (pHu, drip loss, androstenone, marbling, intramuscular fat, texture, and moisture), but no significant differences were detected in sensory analysis. Associations between protein peaks obtained with SELDI-TOF-MS and meat quality traits, mainly water holding capacity, texture and skatole were observed. Of these peaks, a total of 10 peaks from CM10 array and 6 peaks from Q10 array were candidate soluble protein markers for pork loin quality. The developed models explained a limited proportion of the variability, however they point out interesting relationships between protein expression and meat quality

Pitfalls in aminoacid and organic acid analysis: 3-hydroxypropionic aciduria.

Pitfalls in organic acid analysis can originate from inadequate methodology, analytical interferences, in vivo interactions and from pre-analytical conditions which often are unknown to the specialized analytical laboratory. Among the latter, ingested food and additives, metabolites of food processing or medications have to be considered. Bacterial metabolites from the gastrointestinal or urogenital system or formed after sample collection can lead to pitfalls as well. An example of such a patient whose urinary metabolites mimic at first glance inherited propionic aciduria is described.