Topological Descriptors in Modeling the HIV Inhibitory Activity of 2-Aryl-3-pyridyl-thiazolidin-4-ones

The HIV-1 RT inhibitory activity of 2-(2,6-dihalophenyl)-3-(substituted pyridin-2-yl)-thiazolidin-4-ones has been analyzed with different topological descriptors obtained from DRAGON software. Here, simple topological descriptors (TOPO), Galvez topological charge indices (GVZ) and 2D autocorrelation descriptors (2DAUTO) have been found to yield good predictive models for the activity of these compounds. The correlations obtained from the TOPO class descriptors suggest that less extended or compact saturated structural templates would be better for the activity. The participating GVZ class descriptors suggest that they have same degree of influence on the activity. In 2DAUTO class, the large participation of descriptors of lags seven and three indicate the association of activity information with the seven and three centered structural fragments of these compounds. The physicochemical weighting components of these descriptors suggest homogeneous influence of mass, volume, electronegativity and/ or polarizability on the activity.

Soil abiotic factors influence interactions between belowground herbivores and plant roots

Root herbivores are important ecosystem drivers and agricultural pests, and, possibly as a consequence, plants protect their roots using a variety of defensive strategies. One aspect that distinguishes belowground from aboveground plant–insect interactions is that roots are constantly exposed to a set of soil-specific abiotic factors. These factors can profoundly influence root resistance, and, consequently, the outcome of the interaction with belowground feeders. In this review, we synthesize the current literature on the impact of soil moisture, nutrients, and texture on root–herbivore interactions. We show that soil abiotic factors influence the interaction by modulating herbivore abundance and behaviour, root growth and resistance, beneficial microorganisms, as well as natural enemies of the herbivores. We suggest that abiotic heterogeneity may explain the high variability that is often encountered in root–herbivore systems. We also propose that under abiotic stress, the relative fitness value of the roots and the potential negative impact of herbivory increases, which may lead to a higher defensive investment and an increased recruitment of beneficial microorganisms by the plant. At the same time, both root-feeding herbivores and natural enemies are likely to decrease in abundance under extreme environmental conditions, leading to a context- and species-specific impact on plant fitness. Only by using tightly controlled experiments that include soil abiotic heterogeneity will it be possible to understand the impact of root feeders on an ecosystem scale and to develop predictive models for pest occurrence and impact.

Bayesian and survival model for tumor relapse status and disease-specific survival, with applications to breast cancer

Breast cancer is the most common non-skin cancer and the second leading cause of cancer-related death in women in the United States. Studies on ipsilateral breast tumor relapse (IBTR) status and disease-specific survival will help guide clinic treatment and predict patient prognosis.^ After breast conservation therapy, patients with breast cancer may experience breast tumor relapse. This relapse is classified into two distinct types: true local recurrence (TR) and new ipsilateral primary tumor (NP). However, the methods used to classify the relapse types are imperfect and are prone to misclassification. In addition, some observed survival data (e.g., time to relapse and time from relapse to death)are strongly correlated with relapse types. The first part of this dissertation presents a Bayesian approach to (1) modeling the potentially misclassified relapse status and the correlated survival information, (2) estimating the sensitivity and specificity of the diagnostic methods, and (3) quantify the covariate effects on event probabilities. A shared frailty was used to account for the within-subject correlation between survival times. The inference was conducted using a Bayesian framework via Markov Chain Monte Carlo simulation implemented in softwareWinBUGS. Simulation was used to validate the Bayesian method and assess its frequentist properties. The new model has two important innovations: (1) it utilizes the additional survival times correlated with the relapse status to improve the parameter estimation, and (2) it provides tools to address the correlation between the two diagnostic methods conditional to the true relapse types.^ Prediction of patients at highest risk for IBTR after local excision of ductal carcinoma in situ (DCIS) remains a clinical concern. The goals of the second part of this dissertation were to evaluate a published nomogram from Memorial Sloan-Kettering Cancer Center, to determine the risk of IBTR in patients with DCIS treated with local excision, and to determine whether there is a subset of patients at low risk of IBTR. Patients who had undergone local excision from 1990 through 2007 at MD Anderson Cancer Center with a final diagnosis of DCIS (n=794) were included in this part. Clinicopathologic factors and the performance of the Memorial Sloan-Kettering Cancer Center nomogram for prediction of IBTR were assessed for 734 patients with complete data. Nomogram for prediction of 5- and 10-year IBTR probabilities were found to demonstrate imperfect calibration and discrimination, with an area under the receiver operating characteristic curve of .63 and a concordance index of .63. In conclusion, predictive models for IBTR in DCIS patients treated with local excision are imperfect. Our current ability to accurately predict recurrence based on clinical parameters is limited.^ The American Joint Committee on Cancer (AJCC) staging of breast cancer is widely used to determine prognosis, yet survival within each AJCC stage shows wide variation and remains unpredictable. For the third part of this dissertation, biologic markers were hypothesized to be responsible for some of this variation, and the addition of biologic markers to current AJCC staging were examined for possibly provide improved prognostication. The initial cohort included patients treated with surgery as first intervention at MDACC from 1997 to 2006. Cox proportional hazards models were used to create prognostic scoring systems. AJCC pathologic staging parameters and biologic tumor markers were investigated to devise the scoring systems. Surveillance Epidemiology and End Results (SEER) data was used as the external cohort to validate the scoring systems. Binary indicators for pathologic stage (PS), estrogen receptor status (E), and tumor grade (G) were summed to create PS+EG scoring systems devised to predict 5-year patient outcomes. These scoring systems facilitated separation of the study population into more refined subgroups than the current AJCC staging system. The ability of the PS+EG score to stratify outcomes was confirmed in both internal and external validation cohorts. The current study proposes and validates a new staging system by incorporating tumor grade and ER status into current AJCC staging. We recommend that biologic markers be incorporating into revised versions of the AJCC staging system for patients receiving surgery as the first intervention.^ Chapter 1 focuses on developing a Bayesian method to solve misclassified relapse status and application to breast cancer data. Chapter 2 focuses on evaluation of a breast cancer nomogram for predicting risk of IBTR in patients with DCIS after local excision gives the statement of the problem in the clinical research. Chapter 3 focuses on validation of a novel staging system for disease-specific survival in patients with breast cancer treated with surgery as the first intervention. ^

Targeting TRAF6 for Cancer Therapeutical Development

Tumor necrosis factor (TNF)-Receptor Associated Factors (TRAFs) are a family of signal transducer proteins. TRAF6 is a unique member of this family in that it is involved in not only the TNF superfamily, but the toll-like receptor (TLR)/IL-1R (TIR) superfamily. The formation of the complex consisting of Receptor Activator of Nuclear Factor κ B (RANK), with its ligand (RANKL) results in the recruitment of TRAF6, which activates NF-κB, JNK and MAP kinase pathways. TRAF6 is critical in signaling with leading to release of various growth factors in bone, and promotes osteoclastogenesis. TRAF6 has also been implicated as an oncogene in lung cancer and as a target in multiple myeloma. In the hopes of developing small molecule inhibitors of the TRAF6-RANK interaction, multiple steps were carried out. Computational prediction of hot spot residues on the protein-protein interaction of TRAF6 and RANK were examined. Three methods were used: Robetta, KFC2, and HotPoint, each of which uses a different methodology to determine if a residue is a hot spot. These hot spot predictions were considered the basis for resolving the binding site for in silico high-throughput screening using GOLD and the MyriaScreen database of drug/lead-like compounds. Computationally intensive molecular dynamics simulations highlighted the binding mechanism and TRAF6 structural changes upon hit binding. Compounds identified as hits were verified using a GST-pull down assay, comparing inhibition to a RANK decoy peptide. Since many drugs fail due to lack of efficacy and toxicity, predictive models for the evaluation of the LD50 and bioavailability of our TRAF6 hits, and these models can be used towards other drugs and small molecule therapeutics as well. Datasets of compounds and their corresponding bioavailability and LD50 values were curated based, and QSAR models were built using molecular descriptors of these compounds using the k-nearest neighbor (k-NN) method, and quality of these models were cross-validated.

Data quality in trauma transfusion studies and the impact of missing data on predicting massive transfusion

Maximizing data quality may be especially difficult in trauma-related clinical research. Strategies are needed to improve data quality and assess the impact of data quality on clinical predictive models. This study had two objectives. The first was to compare missing data between two multi-center trauma transfusion studies: a retrospective study (RS) using medical chart data with minimal data quality review and the PRospective Observational Multi-center Major Trauma Transfusion (PROMMTT) study with standardized quality assurance. The second objective was to assess the impact of missing data on clinical prediction algorithms by evaluating blood transfusion prediction models using PROMMTT data. RS (2005-06) and PROMMTT (2009-10) investigated trauma patients receiving ≥ 1 unit of red blood cells (RBC) from ten Level I trauma centers. Missing data were compared for 33 variables collected in both studies using mixed effects logistic regression (including random intercepts for study site). Massive transfusion (MT) patients received ≥ 10 RBC units within 24h of admission. Correct classification percentages for three MT prediction models were evaluated using complete case analysis and multiple imputation based on the multivariate normal distribution. A sensitivity analysis for missing data was conducted to estimate the upper and lower bounds of correct classification using assumptions about missing data under best and worst case scenarios. Most variables (17/33=52%) had <1% missing data in RS and PROMMTT. Of the remaining variables, 50% demonstrated less missingness in PROMMTT, 25% had less missingness in RS, and 25% were similar between studies. Missing percentages for MT prediction variables in PROMMTT ranged from 2.2% (heart rate) to 45% (respiratory rate). For variables missing >1%, study site was associated with missingness (all p≤0.021). Survival time predicted missingness for 50% of RS and 60% of PROMMTT variables. MT models complete case proportions ranged from 41% to 88%. Complete case analysis and multiple imputation demonstrated similar correct classification results. Sensitivity analysis upper-lower bound ranges for the three MT models were 59-63%, 36-46%, and 46-58%. Prospective collection of ten-fold more variables with data quality assurance reduced overall missing data. Study site and patient survival were associated with missingness, suggesting that data were not missing completely at random, and complete case analysis may lead to biased results. Evaluating clinical prediction model accuracy may be misleading in the presence of missing data, especially with many predictor variables. The proposed sensitivity analysis estimating correct classification under upper (best case scenario)/lower (worst case scenario) bounds may be more informative than multiple imputation, which provided results similar to complete case analysis.^

Biogenic and lithogenic silica, silicic acid, production of biogenic silica, and irradiance measurements during the Ross Sea Bloom Project 1994-1996

The development of the seasonal phytoplankton bloom in the Ross Sea was studied during two cruises. The first, conducted in November-December 1994, investigated the initiation and rapid growth of the bloom, whereas the second (December 1995-January 1996) concentrated on the bloom's maximum biomass period and the subsequent decline in biomass. Central to the understanding of the controls of growth and the summer decline of the bloom is a quantitative assessment of the growth rate of phytoplankton. Growth rates were estimated over two time scales with different methods. The first estimated daily growth rates from isotropic incorporation under simulated in situ conditions, including 14C, 15N and 32Si uptake measurements combined with estimates of standing stocks of particulate organic carbon, nitrogen and biogenic silica. The second method used daily to weekly changes in biomass at selected locations, with net growth rates being estimated from changes in standing stocks of phytoplankton. In addition, growth rates were estimated in large-volume experiments under optimal irradiances. Growth rates showed distinct temporal patterns. Early in the growing season, short-term estimates suggested that growth rates of in situ assemblages were less than maximum (relative to the temperature-limited maximum) and were likely reduced due to low irradiance regimes encountered under the ice. Growth rates increased thereafter and appeared to reach their maximum as biomass approached the seasonal peak, but decreased markedly in late December. Differences between the major taxonomic groups present were also noted, especially from the isotopic tracer experiments. The haplophyte Phaeocystic antarctica was dominant in 1994 throughout the growing season, and it exhibited the greatest growth rates (mean 0.41/day) during spring. Diatom standing stocks were low early in the growing season, and growth rates averaged 0.100/day. In summer diatoms were more abundant, but their growth rates remained much lower (mean of 0.08/day) than the potential maximum. Understanding growth rate controls is essential to the development of predictive models of the carbon cycle and food webs in Antarctic waters.

Secchi depth maps for the Baltic Sea, 2003-2012, with links to GeoTIFs

Secchi depth is a measure of water transparency. In the Baltic Sea region, Secchi depth maps are used to assess eutrophication and as input for habitat models. Due to their spatial and temporal coverage, satellite data would be the most suitable data source for such maps. But the Baltic Sea's optical properties are so different from the open ocean that globally calibrated standard models suffer from large errors. Regional predictive models that take the Baltic Sea's special optical properties into account are thus needed. This paper tests how accurately generalized linear models (GLMs) and generalized additive models (GAMs) with MODIS/Aqua and auxiliary data as inputs can predict Secchi depth at a regional scale. It uses cross-validation to test the prediction accuracy of hundreds of GAMs and GLMs with up to 5 input variables. A GAM with 3 input variables (chlorophyll a, remote sensing reflectance at 678 nm, and long-term mean salinity) made the most accurate predictions. Tested against field observations not used for model selection and calibration, the best model's mean absolute error (MAE) for daily predictions was 1.07 m (22%), more than 50% lower than for other publicly available Baltic Sea Secchi depth maps. The MAE for predicting monthly averages was 0.86 m (15%). Thus, the proposed model selection process was able to find a regional model with good prediction accuracy. It could be useful to find predictive models for environmental variables other than Secchi depth, using data from other satellite sensors, and for other regions where non-standard remote sensing models are needed for prediction and mapping. Annual and monthly mean Secchi depth maps for 2003-2012 come with this paper as Supplementary materials.

Mineral and oxygen isotopic compositions in a profile through the altered lower oceanic crust, ODP Hole 176-735B

ODP Hole 735B located on the Southwest Indian Ridge at 57°E is an in situ sampled long, continuous section of lower oceanic crust. Oxygen isotope compositions of constituent minerals of Leg 176 gabbros have been measured by UV-laser oxygen isotope microprobe. Together with existing data from Leg 118, a complete oxygen isotope profile through the lower oceanic crust has been obtained. Most clinopyroxenes and olivines have normal mantle values of ~5.5 per mil and ~5.2 per mil, respectively, while plagioclases show slight d18O enrichment relative to its mantle value of 6.1per mil. Down-hole variations of Hole 735B gabbro indicate a downward decreasing d18O profile, with a kink at a depth of about 800 m below sea floor. Above this depth, gabbros are depleted in 18O relative to unaltered basalts, while below ~800 m they show nearly unmodified d18O values. Abundant seawater penetration appears to be limited to the upper part of the lower crust at ODP site 735 (~800 m into the gabbroic layer and ~2-2.5 km into the oceanic crust from the top of pillow basalts). Mass balance calculations show that the lower crust formed under this ultra-slow-spreading ridge has an average d18O value of 5.5 per mil. The whole crust at Site 735 has an overall 18O enrichment with d18O values of 6.0 per mil to 7.8 per mil, depending on the possible variation of the d18O values of the upper pillow basalts and sheeted dykes. The apparent difference in oxygen isotope compositions of ocean crusts formed with different spreading rates has important implications on the buffering of ocean water over geological time, as well as on the oxygen recycling between crust and mantle through subduction. The difference of seawater penetration between fast- and slow-spreading ridges could be related to their particular magmatic-tectonic history during the formation and aging of the crust. However, more analyses on continuous sections through oceanic and ophiolitic crust in different tectonic settings are required to derive any predictive models.

Density of alkyl esters and its mixtures

Biodiesel density is a key parameter in biodiesel simulations and process development. In this work we selected, evaluated and improved two density models, one theoretical (Rackett-Soave) and one empirical (Lapuerta's method) for methanol based biodiesels (FAME) and ethanol based biodiesel (FAEE). For this purpose, biodiesel was produced from vegetable oils (sunflower, rapeseed, soybean, olive, safflower and other two commercial mixtures of vegetable oils) and animal fats (edible and crude pork fat and beef tallow) using both methanol and ethanol for the transesterification reactions, and blended to get 21 FAME and 21 FAEE, reporting their density and detailed composition. Bibliographic data have also been used. The Rackett-Soave method has been improved by the use of a new acentric factor correlation, whereas the parameters of the empirical one are improved by considering a bigger density data bank. Results show that the evaluated models could be used to estimate the biodiesel density with a good grade of accuracy but the performed modifications improve the accuracy of the models: ARD (%) for FAME; 0.33, and FAEE; 0.26, both calculated with the modification of Rackett-Soave method and ARD (%) for FAME; 0.40 calculated with the modification of the Lapuerta's method).

Coral-algae metabolism and diurnal changes in the CO2-carbonate system of bulk sea water

Precise measurements were conducted in continuous flow seawater mesocosms located in full sunlight that compared metabolic response of coral, coral-macroalgae and macroalgae systems over a diurnal cycle. Irradiance controlled net photosynthesis (Pnet), which in turn drove net calcification (Gnet), and altered pH. Pnet exerted the dominant control on [CO3]2- and aragonite saturation state (Omega arag) over the diel cycle. Dark calcification rate decreased after sunset, reaching zero near midnight followed by an increasing rate that peaked at 03:00 h. Changes in Omega arag and pH lagged behind Gnet throughout the daily cycle by two or more hours. The flux rate Pnet was the primary driver of calcification. Daytime coral metabolism rapidly removes dissolved inorganic carbon (DIC) from the bulk seawater and photosynthesis provides the energy that drives Gnet while increasing the bulk water pH. These relationships result in a correlation between Gnet and Omega arag, with Omega arag as the dependent variable. High rates of H+ efflux continued for several hours following mid-day peak Gnet suggesting that corals have difficulty in shedding waste protons as described by the Proton Flux Hypothesis. DIC flux (uptake) followed Pnet and Gnet and dropped off rapidly following peak Pnet and peak Gnet indicating that corals can cope more effectively with the problem of limited DIC supply compared to the problem of eliminating H+. Over a 24 h period the plot of total alkalinity (AT) versus DIC as well as the plot of Gnet versus Omega arag revealed a circular hysteresis pattern over the diel cycle in the coral and coral-algae mesocosms, but not the macroalgae mesocosm. Presence of macroalgae did not change Gnet of the corals, but altered the relationship between Omega arag and Gnet. Predictive models of how future global changes will effect coral growth that are based on oceanic Omega arag must include the influence of future localized Pnet on Gnet and changes in rate of reef carbonate dissolution. The correlation between Omega arag and Gnet over the diel cycle is simply the response of the CO2-carbonate system to increased pH as photosynthesis shifts the equilibria and increases the [CO3]2- relative to the other DIC components of [HCO3]- and [CO2]. Therefore Omega arag closely tracked pH as an effect of changes in Pnet, which also drove changes in Gnet. Measurements of DIC flux and H+ flux are far more useful than concentrations in describing coral metabolism dynamics. Coral reefs are systems that exist in constant disequilibrium with the water column.

Modelización de costas rocosas acantiladas

El retroceso de las costas acantiladas es un fenómeno muy extendido sobre los litorales rocosos expuestos a la incidencia combinada de los procesos marinos y meteorológicos que se dan en la franja costera. Este fenómeno se revela violentamente como movimientos gravitacionales del terreno esporádicos, pudiendo causar pérdidas materiales y/o humanas. Aunque el conocimiento de estos riesgos de erosión resulta de vital importancia para la correcta gestión de la costa, el desarrollo de modelos predictivos se encuentra limitado desde el punto de vista geomorfológico debido a la complejidad e interacción de los procesos de desarrollo espacio-temporal que tienen lugar en la zona costera. Los modelos de predicción publicados son escasos y con importantes inconvenientes: a) extrapolación, extienden la información de registros históricos; b) empíricos, sobre registros históricos estudian la respuesta al cambio de un parámetro; c) estocásticos, determinan la cadencia y magnitud de los eventos futuros extrapolando las distribuciones de probabilidad extraídas de catálogos históricos; d) proceso-respuesta, de estabilidad y propagación del error inexplorada; e) en Ecuaciones en Derivadas Parciales, computacionalmente costosos y poco exactos. La primera parte de esta tesis detalla las principales características de los modelos más recientes de cada tipo y, para los más habitualmente utilizados, se indican sus rangos de aplicación, ventajas e inconvenientes. Finalmente como síntesis de los procesos más relevantes que contemplan los modelos revisados, se presenta un diagrama conceptual de la recesión costera, donde se recogen los procesos más influyentes que deben ser tenidos en cuenta, a la hora de utilizar o crear un modelo de recesión costera con el objetivo de evaluar la peligrosidad (tiempo/frecuencia) del fenómeno a medio-corto plazo. En esta tesis se desarrolla un modelo de proceso-respuesta de retroceso de acantilados costeros que incorpora el comportamiento geomecánico de materiales cuya resistencia a compresión no supere los 5 MPa. El modelo simula la evolución espaciotemporal de un perfil-2D del acantilado que puede estar formado por materiales heterogéneos. Para ello, se acoplan la dinámica marina: nivel medio del mar, cambios en el nivel medio del lago, mareas y oleaje; con la evolución del terreno: erosión, desprendimiento rocoso y formación de talud de derrubios. El modelo en sus diferentes variantes es capaz de incluir el análisis de la estabilidad geomecánica de los materiales, el efecto de los derrubios presentes al pie del acantilado, el efecto del agua subterránea, la playa, el run-up, cambios en el nivel medio del mar o cambios (estacionales o interanuales) en el nivel medio de la masa de agua (lagos). Se ha estudiado el error de discretización del modelo y su propagación en el tiempo a partir de las soluciones exactas para los dos primeros periodos de marea para diferentes aproximaciones numéricas tanto en tiempo como en espacio. Los resultados obtenidos han permitido justificar las elecciones que minimizan el error y los métodos de aproximación más adecuados para su posterior uso en la modelización. El modelo ha sido validado frente a datos reales en la costa de Holderness, Yorkshire, Reino Unido; y en la costa norte del lago Erie, Ontario, Canadá. Los resultados obtenidos presentan un importante avance en los modelos de recesión costera, especialmente en su relación con las condiciones geomecánicas del medio, la influencia del agua subterránea, la verticalización de los perfiles rocosos y su respuesta ante condiciones variables producidas por el cambio climático (por ejemplo, nivel medio del mar, cambios en los niveles de lago, etc.). The recession of coastal cliffs is a widespread phenomenon on the rocky shores that are exposed to the combined incidence of marine and meteorological processes that occur in the shoreline. This phenomenon is revealed violently and occasionally, as gravitational movements of the ground and can cause material or human losses. Although knowledge of the risks of erosion is vital for the proper management of the coast, the development of cliff erosion predictive models is limited by the complex interactions between environmental processes and material properties over a range of temporal and spatial scales. Published prediction models are scarce and present important drawbacks: extrapolation, that extend historical records to the future; empirical, that based on historical records studies the system response against the change in one parameter; stochastic, that represent of cliff behaviour based on assumptions regarding the magnitude and frequency of events in a probabilistic framework based on historical records; process-response, stability and error propagation unexplored; PDE´s, highly computationally expensive and not very accurate. The first part of this thesis describes the main features of the latest models of each type and, for the most commonly used, their ranges of application, advantages and disadvantages are given. Finally as a synthesis of the most relevant processes that include the revised models, a conceptual diagram of coastal recession is presented. This conceptual model includes the most influential processes that must be taken into account when using or creating a model of coastal recession to evaluate the dangerousness (time/frequency) of the phenomenon to medium-short term. A new process-response coastal recession model developed in this thesis has been designed to incorporate the behavioural and mechanical characteristics of coastal cliffs which are composed of with materials whose compressive strength is less than 5 MPa. The model simulates the spatial and temporal evolution of a cliff-2D profile that can consist of heterogeneous materials. To do so, marine dynamics: mean sea level, waves, tides, lake seasonal changes; is coupled with the evolution of land recession: erosion, cliff face failure and associated protective colluvial wedge. The model in its different variants can include analysis of material geomechanical stability, the effect of debris present at the cliff foot, groundwater effects, beach and run-up effects, changes in the mean sea level or changes (seasonal or inter-annual) in the mean lake level. Computational implementation and study of different numerical resolution techniques, in both time and space approximations, and the produced errors are exposed and analysed for the first two tidal periods. The results obtained in the errors analysis allow us to operate the model with a configuration that minimizes the error of the approximation methods. The model is validated through profile evolution assessment at various locations of coastline retreat on the Holderness Coast, Yorkshire, UK and on the north coast of Lake Erie, Ontario, Canada. The results represent an important stepforward in linking material properties to the processes of cliff recession, in considering the effect of groundwater charge and the slope oversteeping and their response to changing conditions caused by climate change (i.e. sea level, changes in lakes levels, etc.).

Análisis de los factores determinantes de la repetición del proceso de compra en el comercio electrónico entre empresas y consumidores finales. Aplicación al mercado español

Las Tecnologías de la Información y la Comunicación en general e Internet en particular han supuesto una revolución en nuestra forma de comunicarnos, relacionarnos, producir, comprar y vender acortando tiempo y distancias entre proveedores y consumidores. A la paulatina penetración del ordenador, los teléfonos inteligentes y la banda ancha fija y/o móvil ha seguido un mayor uso de estas tecnologías entre ciudadanos y empresas. El comercio electrónico empresa–consumidor (B2C) alcanzó en 2010 en España un volumen de 9.114 millones de euros, con un incremento del 17,4% respecto al dato registrado en 2009. Este crecimiento se ha producido por distintos hechos: un incremento en el porcentaje de internautas hasta el 65,1% en 2010 de los cuales han adquirido productos o servicios a través de la Red un 43,1% –1,6 puntos porcentuales más respecto a 2010–. Por otra parte, el gasto medio por comprador ha ascendido a 831€ en 2010, lo que supone un incremento del 10,9% respecto al año anterior. Si segmentamos a los compradores según por su experiencia anterior de compra podemos encontrar dos categorías: el comprador novel –que adquirió por primera vez productos o servicios en 2010– y el comprador constante –aquel que había adquirido productos o servicios en 2010 y al menos una vez en años anteriores–. El 85,8% de los compradores se pueden considerar como compradores constantes: habían comprado en la Red en 2010, pero también lo habían hecho anteriormente. El comprador novel tiene un perfil sociodemográfico de persona joven de entre 15–24 años, con estudios secundarios, de clase social media y media–baja, estudiante no universitario, residente en poblaciones pequeñas y sigue utilizando fórmulas de pago como el contra–reembolso (23,9%). Su gasto medio anual ascendió en 2010 a 449€. El comprador constante, o comprador que ya había comprado en Internet anteriormente, tiene un perfil demográfico distinto: estudios superiores, clase alta, trabajador y residente en grandes ciudades, con un comportamiento maduro en la compra electrónica dada su mayor experiencia –utiliza con mayor intensidad canales exclusivos en Internet que no disponen de tienda presencial–. Su gasto medio duplica al observado en compradores noveles (con una media de 930€ anuales). Por tanto, los compradores constantes suponen una mayoría de los compradores con un gasto medio que dobla al comprador que ha adoptado el medio recientemente. Por consiguiente es de interés estudiar los factores que predicen que un internauta vuelva a adquirir un producto o servicio en la Red. La respuesta a esta pregunta no se ha revelado sencilla. En España, la mayoría de productos y servicios aún se adquieren de manera presencial, con una baja incidencia de las ventas a distancia como la teletienda, la venta por catálogo o la venta a través de Internet. Para dar respuesta a las preguntas planteadas se ha investigado desde distintos puntos de vista: se comenzará con un estudio descriptivo desde el punto de vista de la demanda que trata de caracterizar la situación del comercio electrónico B2C en España, poniendo el foco en las diferencias entre los compradores constantes y los nuevos compradores. Posteriormente, la investigación de modelos de adopción y continuidad en el uso de las tecnologías y de los factores que inciden en dicha continuidad –con especial interés en el comercio electrónico B2C–, permiten afrontar el problema desde la perspectiva de las ecuaciones estructurales pudiendo también extraer conclusiones de tipo práctico. Este trabajo sigue una estructura clásica de investigación científica: en el capítulo 1 se introduce el tema de investigación, continuando con una descripción del estado de situación del comercio electrónico B2C en España utilizando fuentes oficiales (capítulo 2). Posteriormente se desarrolla el marco teórico y el estado del arte de modelos de adopción y de utilización de las tecnologías (capítulo 3) y de los factores principales que inciden en la adopción y continuidad en el uso de las tecnologías (capítulo 4). El capítulo 5 desarrolla las hipótesis de la investigación y plantea los modelos teóricos. Las técnicas estadísticas a utilizar se describen en el capítulo 6, donde también se analizan los resultados empíricos sobre los modelos desarrollados en el capítulo 5. El capítulo 7 expone las principales conclusiones de la investigación, sus limitaciones y propone nuevas líneas de investigación. La primera parte corresponde al capítulo 1, que introduce la investigación justificándola desde un punto de vista teórico y práctico. También se realiza una breve introducción a la teoría del comportamiento del consumidor desde una perspectiva clásica. Se presentan los principales modelos de adopción y se introducen los modelos de continuidad de utilización que se estudiarán más detalladamente en el capítulo 3. En este capítulo se desarrollan los objetivos principales y los objetivos secundarios, se propone el mapa mental de la investigación y se planifican en un cronograma los principales hitos del trabajo. La segunda parte corresponde a los capítulos dos, tres y cuatro. En el capítulo 2 se describe el comercio electrónico B2C en España utilizando fuentes secundarias. Se aborda un diagnóstico del sector de comercio electrónico y su estado de madurez en España. Posteriormente, se analizan las diferencias entre los compradores constantes, principal interés de este trabajo, frente a los compradores noveles, destacando las diferencias de perfiles y usos. Para los dos segmentos se estudian aspectos como el lugar de acceso a la compra, la frecuencia de compra, los medios de pago utilizados o las actitudes hacia la compra. El capítulo 3 comienza desarrollando los principales conceptos sobre la teoría del comportamiento del consumidor, para continuar estudiando los principales modelos de adopción de tecnología existentes, analizando con especial atención su aplicación en comercio electrónico. Posteriormente se analizan los modelos de continuidad en el uso de tecnologías (Teoría de la Confirmación de Expectativas; Teoría de la Justicia), con especial atención de nuevo a su aplicación en el comercio electrónico. Una vez estudiados los principales modelos de adopción y continuidad en el uso de tecnologías, el capítulo 4 analiza los principales factores que se utilizan en los modelos: calidad, valor, factores basados en la confirmación de expectativas –satisfacción, utilidad percibida– y factores específicos en situaciones especiales –por ejemplo, tras una queja– como pueden ser la justicia, las emociones o la confianza. La tercera parte –que corresponde al capítulo 5– desarrolla el diseño de la investigación y la selección muestral de los modelos. En la primera parte del capítulo se enuncian las hipótesis –que van desde lo general a lo particular, utilizando los factores específicos analizados en el capítulo 4– para su posterior estudio y validación en el capítulo 6 utilizando las técnicas estadísticas apropiadas. A partir de las hipótesis, y de los modelos y factores estudiados en los capítulos 3 y 4, se definen y vertebran dos modelos teóricos originales que den respuesta a los retos de investigación planteados en el capítulo 1. En la segunda parte del capítulo se diseña el trabajo empírico de investigación definiendo los siguientes aspectos: alcance geográfico–temporal, tipología de la investigación, carácter y ambiente de la investigación, fuentes primarias y secundarias utilizadas, técnicas de recolección de datos, instrumentos de medida utilizados y características de la muestra utilizada. Los resultados del trabajo de investigación constituyen la cuarta parte de la investigación y se desarrollan en el capítulo 6, que comienza analizando las técnicas estadísticas basadas en Modelos de Ecuaciones Estructurales. Se plantean dos alternativas, modelos confirmatorios correspondientes a Métodos Basados en Covarianzas (MBC) y modelos predictivos. De forma razonada se eligen las técnicas predictivas dada la naturaleza exploratoria de la investigación planteada. La segunda parte del capítulo 6 desarrolla el análisis de los resultados de los modelos de medida y modelos estructurales construidos con indicadores formativos y reflectivos y definidos en el capítulo 4. Para ello se validan, sucesivamente, los modelos de medida y los modelos estructurales teniendo en cuenta los valores umbrales de los parámetros estadísticos necesarios para la validación. La quinta parte corresponde al capítulo 7, que desarrolla las conclusiones basándose en los resultados del capítulo 6, analizando los resultados desde el punto de vista de las aportaciones teóricas y prácticas, obteniendo conclusiones para la gestión de las empresas. A continuación, se describen las limitaciones de la investigación y se proponen nuevas líneas de estudio sobre distintos temas que han ido surgiendo a lo largo del trabajo. Finalmente, la bibliografía recoge todas las referencias utilizadas a lo largo de este trabajo. Palabras clave: comprador constante, modelos de continuidad de uso, continuidad en el uso de tecnologías, comercio electrónico, B2C, adopción de tecnologías, modelos de adopción tecnológica, TAM, TPB, IDT, UTAUT, ECT, intención de continuidad, satisfacción, confianza percibida, justicia, emociones, confirmación de expectativas, calidad, valor, PLS. ABSTRACT Information and Communication Technologies in general, but more specifically those related to the Internet in particular, have changed the way in which we communicate, relate to one another, produce, and buy and sell products, reducing the time and shortening the distance between suppliers and consumers. The steady breakthrough of computers, Smartphones and landline and/or wireless broadband has been greatly reflected in its large scale use by both individuals and businesses. Business–to–consumer (B2C) e–commerce reached a volume of 9,114 million Euros in Spain in 2010, representing a 17.4% increase with respect to the figure in 2009. This growth is due in part to two different facts: an increase in the percentage of web users to 65.1% en 2010, 43.1% of whom have acquired products or services through the Internet– which constitutes 1.6 percentage points higher than 2010. On the other hand, the average spending by individual buyers rose to 831€ en 2010, constituting a 10.9% increase with respect to the previous year. If we select buyers according to whether or not they have previously made some type of purchase, we can divide them into two categories: the novice buyer–who first made online purchases in 2010– and the experienced buyer: who also made purchases in 2010, but had done so previously as well. The socio–demographic profile of the novice buyer is that of a young person between 15–24 years of age, with secondary studies, middle to lower–middle class, and a non–university educated student who resides in smaller towns and continues to use payment methods such as cash on delivery (23.9%). In 2010, their average purchase grew to 449€. The more experienced buyer, or someone who has previously made purchases online, has a different demographic profile: highly educated, upper class, resident and worker in larger cities, who exercises a mature behavior when making online purchases due to their experience– this type of buyer frequently uses exclusive channels on the Internet that don’t have an actual store. His or her average purchase doubles that of the novice buyer (with an average purchase of 930€ annually.) That said, the experienced buyers constitute the majority of buyers with an average purchase that doubles that of novice buyers. It is therefore of interest to study the factors that help to predict whether or not a web user will buy another product or use another service on the Internet. The answer to this question has proven not to be so simple. In Spain, the majority of goods and services are still bought in person, with a low amount of purchases being made through means such as the Home Shopping Network, through catalogues or Internet sales. To answer the questions that have been posed here, an investigation has been conducted which takes into consideration various viewpoints: it will begin with a descriptive study from the perspective of the supply and demand that characterizes the B2C e–commerce situation in Spain, focusing on the differences between experienced buyers and novice buyers. Subsequently, there will be an investigation concerning the technology acceptance and continuity of use of models as well as the factors that have an effect on their continuity of use –with a special focus on B2C electronic commerce–, which allows for a theoretic approach to the problem from the perspective of the structural equations being able to reach practical conclusions. This investigation follows the classic structure for a scientific investigation: the subject of the investigation is introduced (Chapter 1), then the state of the B2C e–commerce in Spain is described citing official sources of information (Chapter 2), the theoretical framework and state of the art of technology acceptance and continuity models are developed further (Chapter 3) and the main factors that affect their acceptance and continuity (Chapter 4). Chapter 5 explains the hypothesis behind the investigation and poses the theoretical models that will be confirmed or rejected partially or completely. In Chapter 6, the technical statistics that will be used are described briefly as well as an analysis of the empirical results of the models put forth in Chapter 5. Chapter 7 explains the main conclusions of the investigation, its limitations and proposes new projects. First part of the project, chapter 1, introduces the investigation, justifying it from a theoretical and practical point of view. It is also a brief introduction to the theory of consumer behavior from a standard perspective. Technology acceptance models are presented and then continuity and repurchase models are introduced, which are studied more in depth in Chapter 3. In this chapter, both the main and the secondary objectives are developed through a mind map and a timetable which highlights the milestones of the project. The second part of the project corresponds to Chapters Two, Three and Four. Chapter 2 describes the B2C e–commerce in Spain from the perspective of its demand, citing secondary official sources. A diagnosis concerning the e–commerce sector and the status of its maturity in Spain is taken on, as well as the barriers and alternative methods of e–commerce. Subsequently, the differences between experienced buyers, which are of particular interest to this project, and novice buyers are analyzed, highlighting the differences between their profiles and their main transactions. In order to study both groups, aspects such as the place of purchase, frequency with which online purchases are made, payment methods used and the attitudes of the purchasers concerning making online purchases are taken into consideration. Chapter 3 begins by developing the main concepts concerning consumer behavior theory in order to continue the study of the main existing acceptance models (among others, TPB, TAM, IDT, UTAUT and other models derived from them) – paying special attention to their application in e–commerce–. Subsequently, the models of technology reuse are analyzed (CDT, ECT; Theory of Justice), focusing again specifically on their application in e–commerce. Once the main technology acceptance and reuse models have been studied, Chapter 4 analyzes the main factors that are used in these models: quality, value, factors based on the contradiction of expectations/failure to meet expectations– satisfaction, perceived usefulness– and specific factors pertaining to special situations– for example, after receiving a complaint justice, emotions or confidence. The third part– which appears in Chapter 5– develops the plan for the investigation and the sample selection for the models that have been designed. In the first section of the Chapter, the hypothesis is presented– beginning with general ideas and then becoming more specific, using the detailed factors that were analyzed in Chapter 4– for its later study and validation in Chapter 6– as well as the corresponding statistical factors. Based on the hypothesis and the models and factors that were studied in Chapters 3 and 4, two original theoretical models are defined and organized in order to answer the questions posed in Chapter 1. In the second part of the Chapter, the empirical investigation is designed, defining the following aspects: geographic–temporal scope, type of investigation, nature and setting of the investigation, primary and secondary sources used, data gathering methods, instruments according to the extent of their use and characteristics of the sample used. The results of the project constitute the fourth part of the investigation and are developed in Chapter 6, which begins analyzing the statistical techniques that are based on the Models of Structural Equations. Two alternatives are put forth: confirmatory models which correspond to Methods Based on Covariance (MBC) and predictive models– Methods Based on Components–. In a well–reasoned manner, the predictive techniques are chosen given the explorative nature of the investigation. The second part of Chapter 6 explains the results of the analysis of the measurement models and structural models built by the formative and reflective indicators defined in Chapter 4. In order to do so, the measurement models and the structural models are validated one by one, while keeping in mind the threshold values of the necessary statistic parameters for their validation. The fifth part corresponds to Chapter 7 which explains the conclusions of the study, basing them on the results found in Chapter 6 and analyzing them from the perspective of the theoretical and practical contributions, and consequently obtaining conclusions for business management. The limitations of the investigation are then described and new research lines about various topics that came up during the project are proposed. Lastly, all of the references that were used during the project are listed in a final bibliography. Key Words: constant buyer, repurchase models, continuity of use of technology, e–commerce, B2C, technology acceptance, technology acceptance models, TAM, TPB, IDT, UTAUT, ECT, intention of repurchase, satisfaction, perceived trust/confidence, justice, feelings, the contradiction of expectations, quality, value, PLS.

Data mining applied to the cognitive rehabilitation of patients with acquired brain injury

Acquired brain injury (ABI) is one of the leading causes of death and disability in the world and is associated with high health care costs as a result of the acute treatment and long term rehabilitation involved. Different algorithms and methods have been proposed to predict the effectiveness of rehabilitation programs. In general, research has focused on predicting the overall improvement of patients with ABI. The purpose of this study is the novel application of data mining (DM) techniques to predict the outcomes of cognitive rehabilitation in patients with ABI. We generate three predictive models that allow us to obtain new knowledge to evaluate and improve the effectiveness of the cognitive rehabilitation process. Decision tree (DT), multilayer perceptron (MLP) and general regression neural network (GRNN) have been used to construct the prediction models. 10-fold cross validation was carried out in order to test the algorithms, using the Institut Guttmann Neurorehabilitation Hospital (IG) patients database. Performance of the models was tested through specificity, sensitivity and accuracy analysis and confusion matrix analysis. The experimental results obtained by DT are clearly superior with a prediction average accuracy of 90.38%, while MLP and GRRN obtained a 78.7% and 75.96%, respectively. This study allows to increase the knowledge about the contributing factors of an ABI patient recovery and to estimate treatment efficacy in individual patients.

Bases metodológicas para un sistema de apoyo a la selección de especies en la restauración de la vegetación

RESUMEN El apoyo a la selección de especies a la restauración de la vegetación en España en los últimos 40 años se ha basado fundamentalmente en modelos de distribución de especies, también llamados modelos de nicho ecológico, que estiman la probabilidad de presencia de las especies en función de las condiciones del medio físico (clima, suelo, etc.). Con esta tesis se ha intentado contribuir a la mejora de la capacidad predictiva de los modelos introduciendo algunas propuestas metodológicas adaptadas a los datos disponibles actualmente en España y enfocadas al uso de los modelos en la selección de especies. No siempre se dispone de datos a una resolución espacial adecuada para la escala de los proyectos de restauración de la vegetación. Sin embrago es habitual contar con datos de baja resolución espacial para casi todas las especies vegetales presentes en España. Se propone un método de recalibración que actualiza un modelo de regresión logística de baja resolución espacial con una nueva muestra de alta resolución espacial. El método permite obtener predicciones de calidad aceptable con muestras relativamente pequeñas (25 presencias de la especie) frente a las muestras mucho mayores (más de 100 presencias) que requería una estrategia de modelización convencional que no usara el modelo previo. La selección del método estadístico puede influir decisivamente en la capacidad predictiva de los modelos y por esa razón la comparación de métodos ha recibido mucha atención en la última década. Los estudios previos consideraban a la regresión logística como un método inferior a técnicas más modernas como las de máxima entropía. Los resultados de la tesis demuestran que esa diferencia observada se debe a que los modelos de máxima entropía incluyen técnicas de regularización y la versión de la regresión logística usada en las comparaciones no. Una vez incorporada la regularización a la regresión logística usando penalización, las diferencias en cuanto a capacidad predictiva desaparecen. La regresión logística penalizada es, por tanto, una alternativa más para el ajuste de modelos de distribución de especies y está a la altura de los métodos modernos con mejor capacidad predictiva como los de máxima entropía. A menudo, los modelos de distribución de especies no incluyen variables relativas al suelo debido a que no es habitual que se disponga de mediciones directas de sus propiedades físicas o químicas. La incorporación de datos de baja resolución espacial proveniente de mapas de suelo nacionales o continentales podría ser una alternativa. Los resultados de esta tesis sugieren que los modelos de distribución de especies de alta resolución espacial mejoran de forma ligera pero estadísticamente significativa su capacidad predictiva cuando se incorporan variables relativas al suelo procedente de mapas de baja resolución espacial. La validación es una de las etapas fundamentales del desarrollo de cualquier modelo empírico como los modelos de distribución de especies. Lo habitual es validar los modelos evaluando su capacidad predictiva especie a especie, es decir, comparando en un conjunto de localidades la presencia o ausencia observada de la especie con las predicciones del modelo. Este tipo de evaluación no responde a una cuestión clave en la restauración de la vegetación ¿cuales son las n especies más idóneas para el lugar a restaurar? Se ha propuesto un método de evaluación de modelos adaptado a esta cuestión que consiste en estimar la capacidad de un conjunto de modelos para discriminar entre las especies presentes y ausentes de un lugar concreto. El método se ha aplicado con éxito a la validación de 188 modelos de distribución de especies leñosas orientados a la selección de especies para la restauración de la vegetación en España. Las mejoras metodológicas propuestas permiten mejorar la capacidad predictiva de los modelos de distribución de especies aplicados a la selección de especies en la restauración de la vegetación y también permiten ampliar el número de especies para las que se puede contar con un modelo que apoye la toma de decisiones. SUMMARY During the last 40 years, decision support tools for plant species selection in ecological restoration in Spain have been based on species distribution models (also called ecological niche models), that estimate the probability of occurrence of the species as a function of environmental predictors (e.g., climate, soil). In this Thesis some methodological improvements are proposed to contribute to a better predictive performance of such models, given the current data available in Spain and focusing in the application of the models to selection of species for ecological restoration. Fine grained species distribution data are required to train models to be used at the scale of the ecological restoration projects, but this kind of data are not always available for every species. On the other hand, coarse grained data are available for almost every species in Spain. A recalibration method is proposed that updates a coarse grained logistic regression model using a new fine grained updating sample. The method allows obtaining acceptable predictive performance with reasonably small updating sample (25 occurrences of the species), in contrast with the much larger samples (more than 100 occurrences) required for a conventional modeling approach that discards the coarse grained data. The choice of the statistical method may have a dramatic effect on model performance, therefore comparisons of methods have received much interest in the last decade. Previous studies have shown a poorer performance of the logistic regression compared to novel methods like maximum entropy models. The results of this Thesis show that the observed difference is caused by the fact that maximum entropy models include regularization techniques and the versions of logistic regression compared do not. Once regularization has been added to the logistic regression using a penalization procedure, the differences in model performance disappear. Therefore, penalized logistic regression may be considered one of the best performing methods to model species distributions. Usually, species distribution models do not consider soil related predictors because direct measurements of the chemical or physical properties are often lacking. The inclusion of coarse grained soil data from national or continental soil maps could be a reasonable alternative. The results of this Thesis suggest that the performance of the models slightly increase after including soil predictors form coarse grained soil maps. Model validation is a key stage of the development of empirical models, such as species distribution models. The usual way of validating is based on the evaluation of model performance for each species separately, i.e., comparing observed species presences or absence to predicted probabilities in a set of sites. This kind of evaluation is not informative for a common question in ecological restoration projects: which n species are the most suitable for the environment of the site to be restored? A method has been proposed to address this question that estimates the ability of a set of models to discriminate among present and absent species in a evaluation site. The method has been successfully applied to the validation of 188 species distribution models used to support decisions on species selection for ecological restoration in Spain. The proposed methodological approaches improve the predictive performance of the predictive models applied to species selection in ecological restoration and increase the number of species for which a model that supports decisions can be fitted.

Propuesta de algoritmos de predicción de glucosa en pacientes diabéticos

La diabetes mellitus es el conjunto de alteraciones provocadas por un defecto en la cantidad de insulina secretada o por un aprovechamiento deficiente de la misma. Es causa directa de complicaciones a corto, medio y largo plazo que disminuyen la calidad y las expectativas de vida de las personas con diabetes. La diabetes mellitus es en la actualidad uno de los problemas más importantes de salud. Ha triplicado su prevalencia en los últimos 20 anos y para el año 2025 se espera que existan casi 300 millones de personas con diabetes. Este aumento de la prevalencia junto con la morbi-mortalidad asociada a sus complicaciones micro y macro-vasculares convierten la diabetes en una carga para los sistemas sanitarios, sus recursos económicos y sus profesionales, haciendo de la enfermedad un problema individual y de salud pública de enormes proporciones. De momento no existe cura a esta enfermedad, de modo que el objetivo terapéutico del tratamiento de la diabetes se centra en la normalización de la glucemia intentando minimizar los eventos de hiper e hipoglucemia y evitando la aparición o al menos retrasando la evolución de las complicaciones vasculares, que constituyen la principal causa de morbi-mortalidad de las personas con diabetes. Un adecuado control diabetológico implica un tratamiento individualizado que considere multitud de factores para cada paciente (edad, actividad física, hábitos alimentarios, presencia de complicaciones asociadas o no a la diabetes, factores culturales, etc.). Sin embargo, a corto plazo, las dos variables más influyentes que el paciente ha de manejar para intervenir sobre su nivel glucémico son la insulina administrada y la dieta. Ambas presentan un retardo entre el momento de su aplicación y el comienzo de su acción, asociado a la absorción de los mismos. Por este motivo la capacidad de predecir la evolución del perfil glucémico en un futuro cercano, ayudara al paciente a tomar las decisiones adecuadas para mantener un buen control de su enfermedad y evitar situaciones de riesgo. Este es el objetivo de la predicción en diabetes: adelantar la evolución del perfil glucémico en un futuro cercano para ayudar al paciente a adaptar su estilo de vida y sus acciones correctoras, con el propósito de que sus niveles de glucemia se aproximen a los de una persona sana, evitando así los síntomas y complicaciones de un mal control. La aparición reciente de los sistemas de monitorización continua de glucosa ha proporcionado nuevas alternativas. La disponibilidad de un registro exhaustivo de las variaciones del perfil glucémico, con un periodo de muestreo de entre uno y cinco minutos, ha favorecido el planteamiento de nuevos modelos que tratan de predecir la glucemia utilizando tan solo las medidas anteriores de glucemia o al menos reduciendo significativamente la información de entrada a los algoritmos. El hecho de requerir menor intervención por parte del paciente, abre nuevas posibilidades de aplicación de los predictores de glucemia, haciéndose viable su uso en tiempo real, como sistemas de ayuda a la decisión, como detectores de situaciones de riesgo o integrados en algoritmos automáticos de control. En esta tesis doctoral se proponen diferentes algoritmos de predicción de glucemia para pacientes con diabetes, basados en la información registrada por un sistema de monitorización continua de glucosa así como incorporando la información de la insulina administrada y la ingesta de carbohidratos. Los algoritmos propuestos han sido evaluados en simulación y utilizando datos de pacientes registrados en diferentes estudios clínicos. Para ello se ha desarrollado una amplia metodología, que trata de caracterizar las prestaciones de los modelos de predicción desde todos los puntos de vista: precisión, retardo, ruido y capacidad de detección de situaciones de riesgo. Se han desarrollado las herramientas de simulación necesarias y se han analizado y preparado las bases de datos de pacientes. También se ha probado uno de los algoritmos propuestos para comprobar la validez de la predicción en tiempo real en un escenario clínico. Se han desarrollado las herramientas que han permitido llevar a cabo el protocolo experimental definido, en el que el paciente consulta la predicción bajo demanda y tiene el control sobre las variables metabólicas. Este experimento ha permitido valorar el impacto sobre el control glucémico del uso de la predicción de glucosa. ABSTRACT Diabetes mellitus is the set of alterations caused by a defect in the amount of secreted insulin or a suboptimal use of insulin. It causes complications in the short, medium and long term that affect the quality of life and reduce the life expectancy of people with diabetes. Diabetes mellitus is currently one of the most important health problems. Prevalence has tripled in the past 20 years and estimations point out that it will affect almost 300 million people by 2025. Due to this increased prevalence, as well as to morbidity and mortality associated with micro- and macrovascular complications, diabetes has become a burden on health systems, their financial resources and their professionals, thus making the disease a major individual and a public health problem. There is currently no cure for this disease, so that the therapeutic goal of diabetes treatment focuses on normalizing blood glucose events. The aim is to minimize hyper- and hypoglycemia and to avoid, or at least to delay, the appearance and development of vascular complications, which are the main cause of morbidity and mortality among people with diabetes. A suitable, individualized and controlled treatment for diabetes involves many factors that need to be considered for each patient: age, physical activity, eating habits, presence of complications related or unrelated to diabetes, cultural factors, etc. However, in the short term, the two most influential variables that the patient has available in order to manage his/her glycemic levels are administered insulin doses and diet. Both suffer from a delay between their time of application and the onset of the action associated with their absorption. Therefore, the ability to predict the evolution of the glycemic profile in the near future could help the patient to make appropriate decisions on how to maintain good control of his/her disease and to avoid risky situations. Hence, the main goal of glucose prediction in diabetes consists of advancing the evolution of glycemic profiles in the near future. This would assist the patient in adapting his/her lifestyle and in taking corrective actions in a way that blood glucose levels approach those of a healthy person, consequently avoiding the symptoms and complications of a poor glucose control. The recent emergence of continuous glucose monitoring systems has provided new alternatives in this field. The availability of continuous records of changes in glycemic profiles (with a sampling period of one or five minutes) has enabled the design of new models which seek to predict blood glucose by using automatically read glucose measurements only (or at least, reducing significantly the data input manually to the algorithms). By requiring less intervention by the patient, new possibilities are open for the application of glucose predictors, making its use feasible in real-time applications, such as: decision support systems, hypo- and hyperglycemia detectors, integration into automated control algorithms, etc. In this thesis, different glucose prediction algorithms are proposed for patients with diabetes. These are based on information recorded by a continuous glucose monitoring system and incorporate information of the administered insulin and carbohydrate intakes. The proposed algorithms have been evaluated in-silico and using patients’ data recorded in different clinical trials. A complete methodology has been developed to characterize the performance of predictive models from all points of view: accuracy, delay, noise and ability to detect hypo- and hyperglycemia. In addition, simulation tools and patient databases have been deployed. One of the proposed algorithms has additionally been evaluated in terms of real-time prediction performance in a clinical scenario in which the patient checked his/her glucose predictions on demand and he/she had control on his/her metabolic variables. This has allowed assessing the impact of using glucose prediction on glycemic control. The tools to carry out the defined experimental protocols were also developed in this thesis.