Sustainable Business models for Open Access Services

Panel at Open Repositories 2014, Helsinki, Finland, June 9-13, 2014

Animal models of anxiety: an ethological perspective

In the field of anxiety research, animal models are used as screening tools in the search for compounds with therapeutic potential and as simulations for research on mechanisms underlying emotional behaviour. However, a solely pharmacological approach to the validation of such tests has resulted in distinct problems with their applicability to systems other than those involving the benzodiazepine/GABAA receptor complex. In this context, recent developments in our understanding of mammalian defensive behaviour have not only prompted the development of new models but also attempts to refine existing ones. The present review focuses on the application of ethological techniques to one of the most widely used animal models of anxiety, the elevated plus-maze paradigm. This fresh approach to an established test has revealed a hitherto unrecognized multidimensionality to plus-maze behaviour and, as it yields comprehensive behavioural profiles, has many advantages over conventional methodology. This assertion is supported by reference to recent work on the effects of diverse manipulations including psychosocial stress, benzodiazepines, GABA receptor ligands, neurosteroids, 5-HT1A receptor ligands, and panicolytic/panicogenic agents. On the basis of this review, it is suggested that other models of anxiety may well benefit from greater attention to behavioural detail

Context-specific independence in graphical models

The theme of this thesis is context-speci c independence in graphical models. Considering a system of stochastic variables it is often the case that the variables are dependent of each other. This can, for instance, be seen by measuring the covariance between a pair of variables. Using graphical models, it is possible to visualize the dependence structure found in a set of stochastic variables. Using ordinary graphical models, such as Markov networks, Bayesian networks, and Gaussian graphical models, the type of dependencies that can be modeled is limited to marginal and conditional (in)dependencies. The models introduced in this thesis enable the graphical representation of context-speci c independencies, i.e. conditional independencies that hold only in a subset of the outcome space of the conditioning variables. In the articles included in this thesis, we introduce several types of graphical models that can represent context-speci c independencies. Models for both discrete variables and continuous variables are considered. A wide range of properties are examined for the introduced models, including identi ability, robustness, scoring, and optimization. In one article, a predictive classi er which utilizes context-speci c independence models is introduced. This classi er clearly demonstrates the potential bene ts of the introduced models. The purpose of the material included in the thesis prior to the articles is to provide the basic theory needed to understand the articles.

The relationship between anxiety and depression in animal models: a study using the forced swimming test and elevated plus-maze

The present study evaluated the correlation between the behavior of mice in the forced swimming test (FST) and in the elevated plus-maze (PM). The effect of the order of the experiments, i.e., the influence of the first test (FST or PM) on mouse behavior in the second test (PM or FST, respectively) was compared to handled animals (HAND). The execution of FST one week before the plus-maze (FST-PM, N = 10), in comparison to mice that were only handled (HAND-PM, N = 10) in week 1, decreased % open entries (HAND-PM: 33.6 ± 2.9; FST-PM: 20.0 ± 3.9; mean ± SEM; P<0.02) and % open time (HAND-PM: 18.9 ± 3.3; FST-PM: 9.0 ± 1.9; P<0.03), suggesting an anxiogenic effect. No significant effect was seen in the number of closed arm entries (FST-PM: 9.5 (7.0-11.0); HAND-PM: 10.0 (4.0-14.5), median (interquartile range); U = 46.5; P>0.10). A prior test in the plus-maze (PM-FST) did not change % immobility time in the FST when compared to the HAND-FST group (HAND-FST: 57.7 ± 3.9; PM-FST: 65.7 ± 3.2; mean ± SEM; P>0.10). Since these data suggest that there is an order effect, the correlation was evaluated separately with each test sequence: FST-PM (N = 20) and PM-FST (N = 18). There was no significant correlation between % immobility time in the FST and plus-maze indexes (% time and entries in open arms) in any test sequence (r: -0.07 to 0.18). These data suggest that mouse behavior in the elevated plus-maze is not related to behavior in the forced swimming test and that a forced swimming test before the plus-maze has an anxiogenic effect even after a one-week interval.

Linguistic models for decision support

Linguistic modelling is a rather new branch of mathematics that is still undergoing rapid development. It is closely related to fuzzy set theory and fuzzy logic, but knowledge and experience from other fields of mathematics, as well as other fields of science including linguistics and behavioral sciences, is also necessary to build appropriate mathematical models. This topic has received considerable attention as it provides tools for mathematical representation of the most common means of human communication - natural language. Adding a natural language level to mathematical models can provide an interface between the mathematical representation of the modelled system and the user of the model - one that is sufficiently easy to use and understand, but yet conveys all the information necessary to avoid misinterpretations. It is, however, not a trivial task and the link between the linguistic and computational level of such models has to be established and maintained properly during the whole modelling process. In this thesis, we focus on the relationship between the linguistic and the mathematical level of decision support models. We discuss several important issues concerning the mathematical representation of meaning of linguistic expressions, their transformation into the language of mathematics and the retranslation of mathematical outputs back into natural language. In the first part of the thesis, our view of the linguistic modelling for decision support is presented and the main guidelines for building linguistic models for real-life decision support that are the basis of our modeling methodology are outlined. From the theoretical point of view, the issues of representation of meaning of linguistic terms, computations with these representations and the retranslation process back into the linguistic level (linguistic approximation) are studied in this part of the thesis. We focus on the reasonability of operations with the meanings of linguistic terms, the correspondence of the linguistic and mathematical level of the models and on proper presentation of appropriate outputs. We also discuss several issues concerning the ethical aspects of decision support - particularly the loss of meaning due to the transformation of mathematical outputs into natural language and the issue or responsibility for the final decisions. In the second part several case studies of real-life problems are presented. These provide background and necessary context and motivation for the mathematical results and models presented in this part. A linguistic decision support model for disaster management is presented here – formulated as a fuzzy linear programming problem and a heuristic solution to it is proposed. Uncertainty of outputs, expert knowledge concerning disaster response practice and the necessity of obtaining outputs that are easy to interpret (and available in very short time) are reflected in the design of the model. Saaty’s analytic hierarchy process (AHP) is considered in two case studies - first in the context of the evaluation of works of art, where a weak consistency condition is introduced and an adaptation of AHP for large matrices of preference intensities is presented. The second AHP case-study deals with the fuzzified version of AHP and its use for evaluation purposes – particularly the integration of peer-review into the evaluation of R&D outputs is considered. In the context of HR management, we present a fuzzy rule based evaluation model (academic faculty evaluation is considered) constructed to provide outputs that do not require linguistic approximation and are easily transformed into graphical information. This is achieved by designing a specific form of fuzzy inference. Finally the last case study is from the area of humanities - psychological diagnostics is considered and a linguistic fuzzy model for the interpretation of outputs of multidimensional questionnaires is suggested. The issue of the quality of data in mathematical classification models is also studied here. A modification of the receiver operating characteristics (ROC) method is presented to reflect variable quality of data instances in the validation set during classifier performance assessment. Twelve publications on which the author participated are appended as a third part of this thesis. These summarize the mathematical results and provide a closer insight into the issues of the practicalapplications that are considered in the second part of the thesis.

Limited-sampling strategy models for estimating the pharmacokinetic parameters of 4-methylaminoantipyrine, an active metabolite of dipyrone

Bioanalytical data from a bioequivalence study were used to develop limited-sampling strategy (LSS) models for estimating the area under the plasma concentration versus time curve (AUC) and the peak plasma concentration (Cmax) of 4-methylaminoantipyrine (MAA), an active metabolite of dipyrone. Twelve healthy adult male volunteers received single 600 mg oral doses of dipyrone in two formulations at a 7-day interval in a randomized, crossover protocol. Plasma concentrations of MAA (N = 336), measured by HPLC, were used to develop LSS models. Linear regression analysis and a "jack-knife" validation procedure revealed that the AUC0-¥ and the Cmax of MAA can be accurately predicted (R²>0.95, bias <1.5%, precision between 3.1 and 8.3%) by LSS models based on two sampling times. Validation tests indicate that the most informative 2-point LSS models developed for one formulation provide good estimates (R²>0.85) of the AUC0-¥ or Cmax for the other formulation. LSS models based on three sampling points (1.5, 4 and 24 h), but using different coefficients for AUC0-¥ and Cmax, predicted the individual values of both parameters for the enrolled volunteers (R²>0.88, bias = -0.65 and -0.37%, precision = 4.3 and 7.4%) as well as for plasma concentration data sets generated by simulation (R²>0.88, bias = -1.9 and 8.5%, precision = 5.2 and 8.7%). Bioequivalence assessment of the dipyrone formulations based on the 90% confidence interval of log-transformed AUC0-¥ and Cmax provided similar results when either the best-estimated or the LSS-derived metrics were used.

Inhibition of the sarcoplasmic reticulum Ca2+ pump with thapsigargin to estimate the contribution of Na+-Ca2+ exchange to ventricular myocyte relaxation

Relaxation in the mammalian ventricle is initiated by Ca2+ removal from the cytosol, which is performed by three main transport systems: sarcoplasmic reticulum Ca2+-ATPase (SR-A), Na+-Ca2+ exchanger (NCX) and the so-called slow mechanisms (sarcolemmal Ca2+-ATPase and mitochondrial Ca2+ uptake). To estimate the relative contribution of each system to twitch relaxation, SR Ca2+ accumulation must be selectively inhibited, usually by the application of high caffeine concentrations. However, caffeine has been reported to often cause changes in membrane potential due to NCX-generated inward current, which compromises the reliability of its use. In the present study, we estimated integrated Ca2+ fluxes carried by SR-A, NCX and slow mechanisms during twitch relaxation, and compared the results when using caffeine application (Cf-NT) and an electrically evoked twitch after inhibition of SR-A with thapsigargin (TG-TW). Ca2+ transients were measured in 20 isolated adult rat ventricular myocytes with indo-1. For transients in which one or more transporters were inhibited, Ca2+ fluxes were estimated from the measured free Ca2+ concentration and myocardial Ca2+ buffering characteristics. NCX-mediated integrated Ca2+ flux was significantly higher with TG-TW than with Cf-NT (12 vs 7 µM), whereas SR-dependent flux was lower with TG-TW (77 vs 81 µM). The relative participations of NCX (12.5 vs 8% with TG-TW and Cf-NT, respectively) and SR-A (85 vs 89.5% with TG-TW and Cf-NT, respectively) in total relaxation-associated Ca2+ flux were also significantly different. We thus propose TG-TW as a reliable alternative to estimate NCX contribution to twitch relaxation in this kind of analysis.

Pressure-assisted cold denaturation of hen egg white lysozyme: the influence of co-solvents probed by hydrogen exchange nuclear magnetic resonance

COSY proton nuclear magnetic resonance was used to measure the exchange rates of amide protons of hen egg white lysozyme (HEWL) in the pressure-assisted cold-denatured state and in the heat-denatured state. After dissolving lysozyme in deuterium oxide buffer, labile protons exchange for deuterons in such a way that exposed protons are substituted rapidly, whereas "protected" protons within structured parts of the protein are substituted slowly. The exchange rates k obs were determined for HEWL under heat treatment (80ºC) and under high pressure conditions at low temperature (3.75 kbar, -13ºC). Moreover, the influence of co-solvents (sorbitol, urea) on the exchange rate was examined under pressure-assisted cold denaturation conditions, and the corresponding protection factors, P, were determined. The exchange kinetics upon heat treatment was found to be a two-step process with initial slow exchange followed by a fast one, showing residual protection in the slow-exchange state and P-factors in the random-coil-like range for the final temperature-denatured state. Addition of sorbitol (500 mM) led to an increase of P-factors for the pressure-assisted cold denatured state, but not for the heat-denatured state. The presence of 2 M urea resulted in a drastic decrease of the P-factors of the pressure-assisted cold denatured state. For both types of co-solvents, the effect they exert appears to be cooperative, i.e., no particular regions within the protein can be identified with significantly diverse changes of P-factors.

Thermodynamic evaluation and modeling of proton and water exchange associated with benzamidine and berenil binding to ß-trypsin

Serine-proteases are involved in vital processes in virtually all species. They are important targets for researchers studying the relationships between protein structure and activity, for the rational design of new pharmaceuticals. Trypsin was used as a model to assess a possible differential contribution of hydration water to the binding of two synthetic inhibitors. Thermodynamic parameters for the association of bovine ß-trypsin (homogeneous material, observed 23,294.4 ± 0.2 Da, theoretical 23,292.5 Da) with the inhibitors benzamidine and berenil at pH 8.0, 25ºC and with 25 mM CaCl2, were determined using isothermal titration calorimetry and the osmotic stress method. The association constant for berenil was about 12 times higher compared to the one for benzamidine (binding constants are K = 596,599 ± 25,057 and 49,513 ± 2,732 M-1, respectively; the number of binding sites is the same for both ligands, N = 0.99 ± 0.05). Apparently the driving force responsible for this large difference of affinity is not due to hydrophobic interactions because the variation in heat capacity (DCp), a characteristic signature of these interactions, was similar in both systems tested (-464.7 ± 23.9 and -477.1 ± 86.8 J K-1 mol-1 for berenil and benzamidine, respectively). The results also indicated that the enzyme has a net gain of about 21 water molecules regardless of the inhibitor tested. It was shown that the difference in affinity could be due to a larger number of interactions between berenil and the enzyme based on computational modeling. The data support the view that pharmaceuticals derived from benzamidine that enable hydrogen bond formation outside the catalytic binding pocket of ß-trypsin may result in more effective inhibitors.

Value at Risk in Foreign Exchange Risk Management

This thesis examines the suitability of VaR in foreign exchange rate risk management from the perspective of a European investor. The suitability of four different VaR models is evaluated in respect to have insight if VaR is a valuable tool in managing foreign exchange rate risk. The models evaluated are historical method, historical bootstrap method, variance-covariance method and Monte Carlo simulation. The data evaluated are divided into emerging and developed market currencies to have more intriguing analysis. The foreign exchange rate data in this thesis is from 31st January 2000 to 30th April 2014. The results show that the previously mentioned VaR models performance in foreign exchange risk management is not to be considered as a single tool in foreign exchange rate risk management. The variance-covariance method and Monte Carlo simulation performs poorest in both currency portfolios. Both historical methods performed better but should also be considered as an additional tool along with other more sophisticated analysis tools. A comparative study of VaR estimates and forward prices is also included in the thesis. The study reveals that regardless of the expensive hedging cost of emerging market currencies the risk captured by VaR is more expensive and thus FX forward hedging is recommended

Deep generative models for facial keypoints detection

A new area of machine learning research called deep learning, has moved machine learning closer to one of its original goals: artificial intelligence and general learning algorithm. The key idea is to pretrain models in completely unsupervised way and finally they can be fine-tuned for the task at hand using supervised learning. In this thesis, a general introduction to deep learning models and algorithms are given and these methods are applied to facial keypoints detection. The task is to predict the positions of 15 keypoints on grayscale face images. Each predicted keypoint is specified by an (x,y) real-valued pair in the space of pixel indices. In experiments, we pretrained deep belief networks (DBN) and finally performed a discriminative fine-tuning. We varied the depth and size of an architecture. We tested both deterministic and sampled hidden activations and the effect of additional unlabeled data on pretraining. The experimental results show that our model provides better results than publicly available benchmarks for the dataset.

Effect of pentoxifylline on lung inflammation and gas exchange in a sepsis-induced acute lung injury model

Experimental models of sepsis-induced pulmonary alterations are important for the study of pathogenesis and for potential intervention therapies. The objective of the present study was to characterize lung dysfunction (low PaO2 and high PaCO2, and increased cellular infiltration, protein extravasation, and malondialdehyde (MDA) production assessed in bronchoalveolar lavage) in a sepsis model consisting of intraperitoneal (ip) injection of Escherichia coli and the protective effects of pentoxifylline (PTX). Male Wistar rats (weighing between 270 and 350 g) were injected ip with 10(7) or 10(9) CFU/100 g body weight or saline and samples were collected 2, 6, 12, and 24 h later (N = 5 each group). PaO2, PaCO2 and pH were measured in blood, and cellular influx, protein extravasation and MDA concentration were measured in bronchoalveolar lavage. In a second set of experiments either PTX or saline was administered 1 h prior to E. coli ip injection (N = 5 each group) and the animals were observed for 6 h. Injection of 10(7) or 10(9) CFU/100 g body weight of E. coli induced acidosis, hypoxemia, and hypercapnia. An increased (P < 0.05) cell influx was observed in bronchoalveolar lavage, with a predominance of neutrophils. Total protein and MDA concentrations were also higher (P < 0.05) in the septic groups compared to control. A higher tumor necrosis factor-alpha (P < 0.05) concentration was also found in these animals. Changes in all parameters were more pronounced with the higher bacterial inoculum. PTX administered prior to sepsis reduced (P < 0.05) most functional alterations. These data show that an E. coli ip inoculum is a good model for the induction of lung dysfunction in sepsis, and suitable for studies of therapeutic interventions.

Effects of pneumonectomy on nitric oxide synthase expression and perivascular edema in the remaining lung of rats

Pneumonectomy is associated with high mortality and high rates of complications. Postpneumonectomy pulmonary edema is one of the leading causes of mortality. Little is known about its etiologic factors and its association with the inflammatory process. The purpose of the present study was to evaluate the role of pneumonectomy as a cause of pulmonary edema and its association with gas exchange, inflammation, nitric oxide synthase (NOS) expression and vasoconstriction. Forty-two non-specific pathogen-free Wistar rats were included in the study. Eleven animals died during or after the procedure, 21 were submitted to left pneumonectomy and 10 to sham operation. These animals were sacrificed after 48 or 72 h. Perivascular pulmonary edema was more intense in pneumonectomized rats at 72 h (P = 0.0131). Neutrophil density was lower after pneumonectomy in both groups (P = 0.0168). There was higher immunohistochemical expression of eNOS in the pneumonectomy group (P = 0.0208), but no statistically significant difference in the expression of iNOS. The lumen-wall ratio and pO2/FiO2 ratio did not differ between the operated and sham groups after pneumonectomy. Left pneumonectomy caused perivascular pulmonary edema with no elevation of immunohistochemical expression of iNOS or neutrophil density, suggesting the absence of correlation with the inflammatory process or oxidative stress. The increased expression of eNOS may suggest an intrinsic production of NO without signs of vascular reactivity.

Animal models of prenatal immune challenge and their contribution to the study of schizophrenia: a systematic review

Prenatal immune challenge (PIC) in pregnant rodents produces offspring with abnormalities in behavior, histology, and gene expression that are reminiscent of schizophrenia and autism. Based on this, the goal of this article was to review the main contributions of PIC models, especially the one using the viral-mimetic particle polyriboinosinic-polyribocytidylic acid (poly-I:C), to the understanding of the etiology, biological basis and treatment of schizophrenia. This systematic review consisted of a search of available web databases (PubMed, SciELO, LILACS, PsycINFO, and ISI Web of Knowledge) for original studies published in the last 10 years (May 2001 to October 2011) concerning animal models of PIC, focusing on those using poly-I:C. The results showed that the PIC model with poly-I:C is able to mimic the prodrome and both the positive and negative/cognitive dimensions of schizophrenia, depending on the specific gestation time window of the immune challenge. The model resembles the neurobiology and etiology of schizophrenia and has good predictive value. In conclusion, this model is a robust tool for the identification of novel molecular targets during prenatal life, adolescence and adulthood that might contribute to the development of preventive and/or treatment strategies (targeting specific symptoms, i.e., positive or negative/cognitive) for this devastating mental disorder, also presenting biosafety as compared to viral infection models. One limitation of this model is the incapacity to model the full spectrum of immune responses normally induced by viral exposure.

Artificial neural network models for predicting 1-year mortality in elderly patients with intertrochanteric fractures in China

The mortality rate of older patients with intertrochanteric fractures has been increasing with the aging of populations in China. The purpose of this study was: 1) to develop an artificial neural network (ANN) using clinical information to predict the 1-year mortality of elderly patients with intertrochanteric fractures, and 2) to compare the ANN's predictive ability with that of logistic regression models. The ANN model was tested against actual outcomes of an intertrochanteric femoral fracture database in China. The ANN model was generated with eight clinical inputs and a single output. ANN's performance was compared with a logistic regression model created with the same inputs in terms of accuracy, sensitivity, specificity, and discriminability. The study population was composed of 2150 patients (679 males and 1471 females): 1432 in the training group and 718 new patients in the testing group. The ANN model that had eight neurons in the hidden layer had the highest accuracies among the four ANN models: 92.46 and 85.79% in both training and testing datasets, respectively. The areas under the receiver operating characteristic curves of the automatically selected ANN model for both datasets were 0.901 (95%CI=0.814-0.988) and 0.869 (95%CI=0.748-0.990), higher than the 0.745 (95%CI=0.612-0.879) and 0.728 (95%CI=0.595-0.862) of the logistic regression model. The ANN model can be used for predicting 1-year mortality in elderly patients with intertrochanteric fractures. It outperformed a logistic regression on multiple performance measures when given the same variables.