Search

In this paper, the performance of new clustering methods such as Neural Gas (NG) and Growing Neural Gas (GNG) is compared with the K-means method for real and simulated data sets. Moreover, a new algorithm called growing K-means, GK, is introduced as the alternative to Neural Gas and Growing Neural Gas. It has small input requirements and is conceptually very simple. The GK leads to nearly optimal values of the cost function, and, contrary to K-means, it is independent of the initial data set partition. The incremental property of GK additionally helps to estimate the number of “natural” clusters in data, i.e., the well-separated groups of objects in the data space.

Mitochondria produce energy through oxidative phosphorylation. A key enzyme in this pathway is F0F1-ATP synthase, catalyzing ATP production from ADP and inorganic phosphate. Recently a subunit of F0F1-ATP synthase, oligomycin sensitivity-conferring protein, was identified as a new estradiol-binding protein. Estradiol could directly modulate mitochondrial ATP synthase activity through this subunit. In addition, intracellular ATP levels play a role in apoptotic death, which is an energy-dependent process requiring functioning mitochondria. Here we examined the effect of 17β -estradiol on F0F1-ATP synthase directly (in permeabilized cells) and in intact osteoclastic FLG 29.1 cells, a model of inducible apoptosis. The baseline F0F1-ATP synthase activity of FLG 29.1 cells was 4.485 nmol/min per mg. Estradiol rapidly inhibited F0F1-ATP synthase activity in the physiological range (half-inhibition concentration, IC50, of 30 nmol/l). With 1 nmol/l of estradiol, the inhibition was already significant (8-10% inhibition, p < 0.01) and with 100 nmol/l residual enzyme activity was only 15% (85% inhibition, p < 0.01). In addition, the effect of estradiol appeared to be directed towards F0F1-ATP synthase, since succinate-sustained respiration, uncoupled from the electron transport chain, was unaffected by estradiol. We assayed F0F1-ATP synthase activity in FLG 29.1 cells during inducible apoptosis. No significant difference of ATP synthesis was detected in apoptotic cells versus controls. In conclusion, we showed a new non-genomic effect of estradiol on a key mitochondrial enzyme, which thereby directly modulates cellular energy metabolism.

The main principles and the algorithm of a density-based clustering approach, OPTICS, are described, and its unique properties, such as the ability to reveal clusters of arbitrary shapes and different densities, are illustrated on simulated and real spectral and chromatographic data sets. A “reachability plot” visualizing density fluctuations of data in multivariate space and a “color map” relating the original and/or descriptive features with data clustering allow a deeper insight into the data structure and its interpretation in chemical terms.

A 2(10-5) fractional factorial model was used to investigate the influence of 10 process variables in broth microdilution susceptibility tests with itraconazole against eight isolates of Candida species and six isolates of filamentous fungi in two growth media. An analysis of variance (ANOVA) indicated that glucose concentration and incubation time both significantly influenced control turbidity optical density (OD) values for most of the Candida spp. isolates, while incubation in >10% CO(2) versus ambient air, incubation temperature and inoculum size significantly influenced these OD values for about half of the yeast isolates. Control OD values for the mould isolates were most influenced by incubation time and temperature, and by occlusion of the wells with an adhesive sticker. Three statistical approaches, ANOVA, rank transformation and Mann-Whitney U-test, were used to assess the influence of the variable combinations on MIC, determined with a 50% growth reduction end-point. Incubation temperature and time, glucose concentration and inoculum size were the variables that most often affected susceptibility results to the level of statistical significance; however, the supplier of RPMI 1640 medium, the use of adhesive stickers and the atmosphere of incubation significantly influenced the MIC for some isolates. The medium used to prepare the test inoculum, the solvent used to prepare the stock solution and the shape of the microdilution plate wells significantly affected outcome, but only sporadically. A principal component analysis of the data matrix confirmed this order of relative influence of the test variables on the MIC. Since each fungal isolate responded differently to combinations of process variables in the test, we conclude that any unified method for antifungal susceptibility determination represents a compromise, rather than an idealized system.

OBJECTIVE: The cyclic AMP (cAMP) cascade is the main regulatory pathway in thyrocytes. Whilst activating mutations in the TSH receptor or in the Gs alpha-subunit, which increase cAMP levels, have been shown to be responsible for 80% of the autonomous adenomas, no such mutations have been observed in the other types of thyroid tumors, suggesting that other mechanisms exist. The discovery of Epac ('exchange nucleotide protein directly activated by cAMP'), a novel cAMP-binding protein, which is strongly expressed in the thyroid, raised the possibility of a role for this protein in the generation of the unexplained cold thyroid follicular adenomas. Thus, we investigated whether activating mutations in either Epac or Rap (the downstream target of Epac) could be responsible for the generation of these thyroid nodules. DESIGN: Epac and Rap1 (Rap1A and Rap1B) cDNAs were sequenced in 10 patients. The sequencing of the cDNAs was realized on both strands in the cold nodule and the juxtanodular tissue of each patient. RESULTS: No mutations in either Epac or Rap1 cDNAs were found. For five patients, a polymorphism in Epac at codon 332 (Gly--Ser) was observed. CONCLUSIONS: In this report, we show that the cAMP--Epac--Rap1 signaling pathway in the thyroid gland does not play a major role in the generation of cold thyroid follicular adenomas, since no mutations in either Epac or Rap1 could be observed in the 10 nodules studied.

Studies on the nutritional content of foodstuffs often lead to large or complex data sets. The situation described here is complex in the sense that the table has a three-way structure (elemental content×type of vegetable×cooking style) and the question is to find the main trends in this data. The use of three-way principal component analysis (PCA) is proposed for this purpose and compared with the use of the better known two-way PCA. How vegetables can be classified according to their elemental content and the effect of cooking style on that content are shown. It is concluded from this example that three-way PCA can be a useful tool in the analysis of nutritional tables.

The environmetrical analysis of the benthic organisms data set from a Black sea region has revealed new information concerning the chemical content and the bioindicating abilities of polychaeta (Melina palmata), Crustacea (Aspendopsis ostroumovi)and molluscs (Mytilus gallo‐provincialis).The application of various multivariate statistical approaches like cluster and principal component analysis, linear regression and partial least square modeling, source apportioning makes it possible to understand in a better way the properties of the benthic organism as collectors of pollutants in a total and a more specific mode. It is shown that heavily polluted coastal zones are indicated in the same way by all benthic species but some specificity could be detected when moderately polluted zones are considered. In this case polychaetaaccumulated preferably Co, Cr, Cu and Pb Crustacea‐ As, Cd and Ni and molluscs‐ Zn to a limited extent.PCA identified three latent factors (“anthropogenic flotation”;, “anthropogenic galvanic”; and “naturally occurring") which explain about 65% of the total variance of the system and determine the data set structure. The source apportioning on the absolute principal component scores proved that none of the metals is quantitatively linked with only one anthropogenic or natural source.The linear regression and PLS models have indicated that a reliable prognosis of the pollution on some naturally occurring chemical components (e.g. linear regression on Zn for Cr and Ni) or combination of them (PLS modeling on Mn/Zn or on Mn/Zn/Fe for the rest of the pollutants) could be achieved.

AbstractThe present study deals with the application of different multivariate statistical approaches to a pot water data set from Athens, Greece. The set represents analytical monitoring data for 16 potable water parameters (turbidity, pH, free acidity, total hardness, chloride, sulfate, nitrate, nitrite, calcium, magnesium, sodium, chemical oxygen demand, dry residue, conductivity, phosphate and silicon dioxide) measured in samples collected at 17 different sampling sites (3 sites of the water purification complex near Athens, 3 sites of lake system Iliki, 5 sites of lake system Marathonas and 6 sites of lake system Mornos) during 6 bimonths intervals in 1998. Both unsupervised and supervised environmetrical methods were used for data treatment: two‐ and multi‐way principal component analysis (PCA and MPCA‐Tucker 3 model) and CART (partitioning tree), respectively. It has been found that “rural”; and “urban”; conditional factors are responsible for the site spatial separation and “salinity”; and “turbidity”; latent factors ‐ for the data structuring. Additionally, a separation in “spring/summer”; and “autumn/winter”; factors is observed determining the seasonal behavior of the water sources. A simple set of discriminant variables (turbidity, total hardness, conductivity, free acidity) could be found making possible an easy separation between water sources and purified pot water samples. It may be stated that the quality of the water in the lake systems Iliki, Marathonas and Mornos with respect to inorganic chemical content and physicochemical properties is quite similar to the quality of the purified pot water delivered from the purification complex of the City of Athens.

We derive a new general concentration‐of‐measure inequality. The concentration inequality applies, among others, to configuration functions as defined by Talagrand and also to combinatorial entropies such as the logarithm of the number of increasing subsequences in a random permutation and to Vapnik‐Chervonenkis (VC) entropies. The results find direct applications in statistical learning theory, substantiating the possibility to use the empirical VC entropy in penalization techniques. © 2000 John Wiley & Sons, Inc. Random Struct. Alg., 16: 277–292, 2000

The present study compares the performance of different multivariate calibration techniques applied to four near-infrared data sets when test samples are well within the calibration domain. Three types of problems are discussed: the nonlinear calibration, the calibration using heterogeneous data sets, and the calibration in the presence of irrelevant information in the set of predictors. Recommendations are derived from the comparison, which should help to guide a nonchemometrician through the selection of an appropriate calibration method for a particular type of calibration data. A flexible methodology is proposed to allow selection of an appropriate calibration technique for a given calibration problem.

The orthogonal projection approach (OPA), principal component analysis (PCA), and multivariate curve resolution alternating least-squares (MCR-ALS) are applied to detect the end point of an organic synthetic process monitored on-line by mid-infrared spectroscopy. The OPA dissimilarity criterion and the PCA plot are useful to follow the changes in the on-line measured spectra and to detect regions where a steady state of the synthesis is obtained. MCR-ALS is applied to obtain semi-quantitative information during the synthesis by analyzing the steady-state regions simultaneously. The concentrations of the components in the different steady-state regions are calculated with respect to their concentration in a reference steady-state region, chosen from the OPA dissimilarity and PCA plots.

The effect of model updating on the identification of a pharmaceutical excipient based on its near-infrared (NIR) spectra has been investigated. A pragmatic updating approach, consisting of adding stepwise newly available samples to the training set and rebuilding the classification model, was applied. Its performance is compared for three pattern recognition methods: the wavelength distance method, the Mahalanobis distance method, and the SIMCA (soft independent modeling of class analogy) residual variance method. For the wavelength distance method, the updating approach is straightforward. In the case of the multivariate classification methods, which are based on a certain number of significant principal components (PCs), the selection of the number of PCs included in the model must be performed with care, as this number has a major impact on the classification results.