Stochastic modeling of bacterial protein domains distribution to predict horizontal gene transfer

In questo elaborato, abbiamo tentato di modellizzare i processi che regolano la presenza dei domini proteici. I domini proteici studiati in questa tesi sono stati ottenuti dai genomi batterici disponibili nei data base pubblici (principalmente dal National Centre for Biotechnology Information: NCBI) tramite una procedura di simulazione computazionale. Ci siamo concentrati su organismi batterici in quanto in essi la presenza di geni trasmessi orizzontalmente, ossia che parte del materiale genetico non provenga dai genitori, e assodato che sia presente in una maggiore percentuale rispetto agli organismi più evoluti. Il modello usato si basa sui processi stocastici di nascita e morte, con l'aggiunta di un parametro di migrazione, usato anche nella descrizione dell'abbondanza relativa delle specie in ambito delle biodiversità ecologiche. Le relazioni tra i parametri, calcolati come migliori stime di una distribuzione binomiale negativa rinormalizzata e adattata agli istogrammi sperimentali, ci induce ad ipotizzare che le famiglie batteriche caratterizzate da un basso valore numerico del parametro di immigrazione abbiano contrastato questo deficit con un elevato valore del tasso di nascita. Al contrario, ipotizziamo che le famiglie con un tasso di nascita relativamente basso si siano adattate, e in conseguenza, mostrano un elevato valore del parametro di migrazione. Inoltre riteniamo che il parametro di migrazione sia direttamente proporzionale alla quantità di trasferimento genico orizzontale effettuato dalla famiglia batterica.

Predicting adverse events in children with fever and chemotherapy-induced neutropenia: the prospective multicenter SPOG 2003 FN study

PURPOSE To develop a score predicting the risk of adverse events (AEs) in pediatric patients with cancer who experience fever and neutropenia (FN) and to evaluate its performance. PATIENTS AND METHODS Pediatric patients with cancer presenting with FN induced by nonmyeloablative chemotherapy were observed in a prospective multicenter study. A score predicting the risk of future AEs (ie, serious medical complication, microbiologically defined infection, radiologically confirmed pneumonia) was developed from a multivariate mixed logistic regression model. Its cross-validated predictive performance was compared with that of published risk prediction rules. Results An AE was reported in 122 (29%) of 423 FN episodes. In 57 episodes (13%), the first AE was known only after reassessment after 8 to 24 hours of inpatient management. Predicting AE at reassessment was better than prediction at presentation with FN. A differential leukocyte count did not increase the predictive performance. The score predicting future AE in 358 episodes without known AE at reassessment used the following four variables: preceding chemotherapy more intensive than acute lymphoblastic leukemia maintenance (weight = 4), hemoglobin > or = 90 g/L (weight = 5), leukocyte count less than 0.3 G/L (weight = 3), and platelet count less than 50 G/L (weight = 3). A score (sum of weights) > or = 9 predicted future AEs. The cross-validated performance of this score exceeded the performance of published risk prediction rules. At an overall sensitivity of 92%, 35% of the episodes were classified as low risk, with a specificity of 45% and a negative predictive value of 93%. CONCLUSION This score, based on four routinely accessible characteristics, accurately identifies pediatric patients with cancer with FN at risk for AEs after reassessment.

A prediction model for assessing residential radon concentration in Switzerland

Indoor radon is regularly measured in Switzerland. However, a nationwide model to predict residential radon levels has not been developed. The aim of this study was to develop a prediction model to assess indoor radon concentrations in Switzerland. The model was based on 44,631 measurements from the nationwide Swiss radon database collected between 1994 and 2004. Of these, 80% randomly selected measurements were used for model development and the remaining 20% for an independent model validation. A multivariable log-linear regression model was fitted and relevant predictors selected according to evidence from the literature, the adjusted R², the Akaike's information criterion (AIC), and the Bayesian information criterion (BIC). The prediction model was evaluated by calculating Spearman rank correlation between measured and predicted values. Additionally, the predicted values were categorised into three categories (50th, 50th-90th and 90th percentile) and compared with measured categories using a weighted Kappa statistic. The most relevant predictors for indoor radon levels were tectonic units and year of construction of the building, followed by soil texture, degree of urbanisation, floor of the building where the measurement was taken and housing type (P-values <0.001 for all). Mean predicted radon values (geometric mean) were 66 Bq/m³ (interquartile range 40-111 Bq/m³) in the lowest exposure category, 126 Bq/m³ (69-215 Bq/m³) in the medium category, and 219 Bq/m³ (108-427 Bq/m³) in the highest category. Spearman correlation between predictions and measurements was 0.45 (95%-CI: 0.44; 0.46) for the development dataset and 0.44 (95%-CI: 0.42; 0.46) for the validation dataset. Kappa coefficients were 0.31 for the development and 0.30 for the validation dataset, respectively. The model explained 20% overall variability (adjusted R²). In conclusion, this residential radon prediction model, based on a large number of measurements, was demonstrated to be robust through validation with an independent dataset. The model is appropriate for predicting radon level exposure of the Swiss population in epidemiological research. Nevertheless, some exposure misclassification and regression to the mean is unavoidable and should be taken into account in future applications of the model.

Model Evaluation Based on the Distribution of Estimated Absolute Prediction Error

The construction of a reliable, practically useful prediction rule for future response is heavily dependent on the "adequacy" of the fitted regression model. In this article, we consider the absolute prediction error, the expected value of the absolute difference between the future and predicted responses, as the model evaluation criterion. This prediction error is easier to interpret than the average squared error and is equivalent to the mis-classification error for the binary outcome. We show that the distributions of the apparent error and its cross-validation counterparts are approximately normal even under a misspecified fitted model. When the prediction rule is "unsmooth", the variance of the above normal distribution can be estimated well via a perturbation-resampling method. We also show how to approximate the distribution of the difference of the estimated prediction errors from two competing models. With two real examples, we demonstrate that the resulting interval estimates for prediction errors provide much more information about model adequacy than the point estimates alone.

A Diagnostic Test for the Mixing Distribution in a Generalised Linear Mixed Model

We introduce a diagnostic test for the mixing distribution in a generalised linear mixed model. The test is based on the difference between the marginal maximum likelihood and conditional maximum likelihood estimates of a subset of the fixed effects in the model. We derive the asymptotic variance of this difference, and propose a test statistic that has a limiting chi-square distribution under the null hypothesis that the mixing distribution is correctly specified. For the important special case of the logistic regression model with random intercepts, we evaluate via simulation the power of the test in finite samples under several alternative distributional forms for the mixing distribution. We illustrate the method by applying it to data from a clinical trial investigating the effects of hormonal contraceptives in women.

Checking Assumptions in Latent Class Regression Models via a Markov Chain Monte Carlo Estimation Approach: An Application to Depression and Socio-Economic Status

Latent class regression models are useful tools for assessing associations between covariates and latent variables. However, evaluation of key model assumptions cannot be performed using methods from standard regression models due to the unobserved nature of latent outcome variables. This paper presents graphical diagnostic tools to evaluate whether or not latent class regression models adhere to standard assumptions of the model: conditional independence and non-differential measurement. An integral part of these methods is the use of a Markov Chain Monte Carlo estimation procedure. Unlike standard maximum likelihood implementations for latent class regression model estimation, the MCMC approach allows us to calculate posterior distributions and point estimates of any functions of parameters. It is this convenience that allows us to provide the diagnostic methods that we introduce. As a motivating example we present an analysis focusing on the association between depression and socioeconomic status, using data from the Epidemiologic Catchment Area study. We consider a latent class regression analysis investigating the association between depression and socioeconomic status measures, where the latent variable depression is regressed on education and income indicators, in addition to age, gender, and marital status variables. While the fitted latent class regression model yields interesting results, the model parameters are found to be invalid due to the violation of model assumptions. The violation of these assumptions is clearly identified by the presented diagnostic plots. These methods can be applied to standard latent class and latent class regression models, and the general principle can be extended to evaluate model assumptions in other types of models.

Towards greater accuracy in individual-tree mortality regression

Background mortality is an essential component of any forest growth and yield model. Forecasts of mortality contribute largely to the variability and accuracy of model predictions at the tree, stand and forest level. In the present study, I implement and evaluate state-of-the-art techniques to increase the accuracy of individual tree mortality models, similar to those used in many of the current variants of the Forest Vegetation Simulator, using data from North Idaho and Montana. The first technique addresses methods to correct for bias induced by measurement error typically present in competition variables. The second implements survival regression and evaluates its performance against the traditional logistic regression approach. I selected the regression calibration (RC) algorithm as a good candidate for addressing the measurement error problem. Two logistic regression models for each species were fitted, one ignoring the measurement error, which is the “naïve” approach, and the other applying RC. The models fitted with RC outperformed the naïve models in terms of discrimination when the competition variable was found to be statistically significant. The effect of RC was more obvious where measurement error variance was large and for more shade-intolerant species. The process of model fitting and variable selection revealed that past emphasis on DBH as a predictor variable for mortality, while producing models with strong metrics of fit, may make models less generalizable. The evaluation of the error variance estimator developed by Stage and Wykoff (1998), and core to the implementation of RC, in different spatial patterns and diameter distributions, revealed that the Stage and Wykoff estimate notably overestimated the true variance in all simulated stands, but those that are clustered. Results show a systematic bias even when all the assumptions made by the authors are guaranteed. I argue that this is the result of the Poisson-based estimate ignoring the overlapping area of potential plots around a tree. Effects, especially in the application phase, of the variance estimate justify suggested future efforts of improving the accuracy of the variance estimate. The second technique implemented and evaluated is a survival regression model that accounts for the time dependent nature of variables, such as diameter and competition variables, and the interval-censored nature of data collected from remeasured plots. The performance of the model is compared with the traditional logistic regression model as a tool to predict individual tree mortality. Validation of both approaches shows that the survival regression approach discriminates better between dead and alive trees for all species. In conclusion, I showed that the proposed techniques do increase the accuracy of individual tree mortality models, and are a promising first step towards the next generation of background mortality models. I have also identified the next steps to undertake in order to advance mortality models further.

Heart rate response determines long term exercise capacity after heart transplantation

BACKGROUND: Exercise capacity after heart transplantation (HTx) remains limited despite normal left ventricular systolic function of the allograft. Various clinical and haemodynamic parameters are predictive of exercise capacity following HTx. However, the predictive significance of chronotropic competence has not been demonstrated unequivocally despite its immediate relevance for cardiac output. AIMS: This study assesses the predictive value of various clinical and haemodynamic parameters for exercise capacity in HTx recipients with complete chronotropic competence evolving within the first 6 postoperative months. METHODS: 51 patients were enrolled in this exercise study. Patients were included when at least >6 months after HTx and without negative chronotropic medication or factors limiting exercise capacity such as significant transplant vasculopathy or allograft rejection. Clinical parameters were obtained by chart review, haemodynamic parameters from current cardiac catheterisation, and exercise capacity was assessed by treadmill stress testing. A stepwise multiple regression model analysed the proportion of the variance explained by the predictive parameters. RESULTS: The mean age of these 51 HTx recipients was 55.4 +/- 13.2 yrs on inclusion, 42 pts were male and the mean time interval after cardiac transplantation was 5.1 +/- 2.8 yrs. Five independent predictors explained 47.5% of the variance observed for peak exercise capacity (adjusted R2 = 0.475). In detail, heart rate response explained 31.6%, male gender 5.2%, age 4.1%, pulmonary vascular resistance 3.7%, and body-mass index 2.9%. CONCLUSION: Heart rate response is one of the most important predictors of exercise capacity in HTx recipients with complete chronotropic competence and without relevant transplant vasculopathy or acute allograft rejection.

Infection with human herpesvirus 8 and transplant-associated gammopathy

BACKGROUND: The role of human herpesvirus (HHV)-8 in the pathogenesis of multiple myeloma and its pre-malignant state of monoclonal gammopathy is unclear. HHV-8 is transmitted by organ transplantation, representing a unique model with which to investigate primary HHV-8 infection. METHODS: The authors studied the incidence of clonal gammopathy in renal transplant recipients and correlated it with previous and recent HHV-8 infection. RESULTS: Clonal gammopathy was observed in 31 of 162 (19%) HHV-8-seronegative patients, in 5 of 17 (29%) HHV-8-seropositive patients, and in 9 of 24 (38%) HHV-8 seroconverters within 5 years after transplantation. Gammopathy was often transient, and no progression to myeloma was observed. Two patients with persistent gammopathy developed B-cell lymphoma. In a logistic regression model, HHV-8 serostatus of the graft recipient was significantly associated with subsequent development of gammopathy, with a relative risk (RR) of 1.9 and a 95% confidence interval (CI) of 0.5 to 6.4 for an HHV-8-seropositive recipient and an RR of 2.9 and a 95% CI of 1.01 to 8.0 for seroconverters as compared with baseline (HHV-8 seronegative). Other significant variables were cytomegalovirus (CMV) serostatus and the intensity of immunosuppression (RR of 10.4 and 95% CI of 2.6-41.7 for a CMV-negative recipient with a CMV-positive donor vs. a CMV-negative recipient with a CMV-negative donor and RR of 17.6 and 95% CI of 2.0-150.8 if OKT3 was used vs. no use of antilymphocytic substances). CONCLUSIONS: Transplant recipients with HHV-8 infection are more likely to develop clonal gammopathy. However, this risk is much lower than the risk conferred by CMV infection and antilymphocytic therapy, arguing against a major role of HHV-8 infection in the pathogenesis of clonal plasma cell proliferation.

Effects of Ownership on Hospital Efficiency in Germany

The objective of our study was to evaluate the efficiency of public, private for-profit, and private non-profit hospitals in Germany. First, bootstrapped data envelopment analysis (DEA) was used to evaluate the efficiency of a panel (n = 1,046) of public, private for-profit, and private non-profit hospitals between 2002 and 2006. This was followed by a second-step truncated linear regression model with bootstrapped DEA efficiency scores as dependent variable. The results show that public hospitals performed significantly better than their private for-profit and non-profit counterparts. In addition, we found a significant positive association between hospital size and efficiency, and that competitive pressure had a significant negative impact on hospital efficiency.

Expanding the cone location and magnitude index to include corneal thickness and posterior surface information for the detection of keratoconus

PURPOSE To extend the capabilities of the Cone Location and Magnitude Index algorithm to include a combination of topographic information from the anterior and posterior corneal surfaces and corneal thickness measurements to further improve our ability to correctly identify keratoconus using this new index: ConeLocationMagnitudeIndex_X. DESIGN Retrospective case-control study. METHODS Three independent data sets were analyzed: 1 development and 2 validation. The AnteriorCornealPower index was calculated to stratify the keratoconus data from mild to severe. The ConeLocationMagnitudeIndex algorithm was applied to all tomography data collected using a dual Scheimpflug-Placido-based tomographer. The ConeLocationMagnitudeIndex_X formula, resulting from analysis of the Development set, was used to determine the logistic regression model that best separates keratoconus from normal and was applied to all data sets to calculate PercentProbabilityKeratoconus_X. The sensitivity/specificity of PercentProbabilityKeratoconus_X was compared with the original PercentProbabilityKeratoconus, which only uses anterior axial data. RESULTS The AnteriorCornealPower severity distribution for the combined data sets are 136 mild, 12 moderate, and 7 severe. The logistic regression model generated for ConeLocationMagnitudeIndex_X produces complete separation for the Development set. Validation Set 1 has 1 false-negative and Validation Set 2 has 1 false-positive. The overall sensitivity/specificity results for the logistic model produced using the ConeLocationMagnitudeIndex_X algorithm are 99.4% and 99.6%, respectively. The overall sensitivity/specificity results for using the original ConeLocationMagnitudeIndex algorithm are 89.2% and 98.8%, respectively. CONCLUSIONS ConeLocationMagnitudeIndex_X provides a robust index that can detect the presence or absence of a keratoconic pattern in corneal tomography maps with improved sensitivity/specificity from the original anterior surface-only ConeLocationMagnitudeIndex algorithm.

Plasma DNA is Elevated in Patients with Deep Vein Thrombosis

OBJECTIVE To investigate if plasma DNA is elevated in patients with deep vein thrombosis (DVT) and to determine whether there is a correlation with other biomarkers of DVT. BACKGROUND Leukocytes release DNA to form extracellular traps (ETs), which have recently been linked to experimental DVT. In baboons and mice, extracellular DNA co-localized with von Willebrand factor (VWF) in the thrombus and DNA appeared in circulation at the time of thrombus formation. ETs have not been associated with clinical DVT. SETTING From December 2008 to August 2010, patients were screened through the University of Michigan Diagnostic Vascular Unit and were divided into three distinct groups: 1) the DVT positive group, consisting of patients who were symptomatic for DVT, which was confirmed by compression duplex ultrasound (n=47); 2) the DVT negative group, consisting of patients that present with swelling and leg pain but had a negative compression duplex ultrasound, (n=28); and 3) a control group of healthy non-pregnant volunteers without signs or symptoms of active or previous DVT (n=19). Patients were excluded if they were less than 18 years of age, unwillingness to consent, pregnant, on an anticoagulant therapy, or diagnosed with isolated calf vein thrombosis. METHODS Blood was collected for circulating DNA, CRP, D-dimer, VWF activity, myeloperoxidase (MPO), ADAMTS13 and VWF. The Wells score for a patient's risk of DVT was assessed. The Receiver Operating Characteristic (ROC) curve was generated to determine the strength of the relationship between circulating DNA levels and the presence of DVT. A Spearman correlation was performed to determine the relationship between the DNA levels and the biomarkers and the Wells score. Additionally the ratio of ADAMTS13/VWF was assessed. RESULTS Our results showed that circulating DNA (a surrogate marker for NETs) was significantly elevated in DVT patients, compared to both DVT negative patients (57.7±6.3 vs. 17.9±3.5ng/mL, P<.01) and controls (57.7±6.3 vs. 23.9±2.1ng/mL, P<.01). There was a strong positive correlation with CRP (P<.01), D-dimer (P<.01), VWF (P<.01), Wells score (P<.01) and myeloperoxidase (MPO) (P<.01), along with a strong negative correlation with ADAMTS13 (P<.01) and the ADAMTS13/VWF ratio. The logistic regression model showed a strong association between plasma DNA and the presence of DVT (ROC curve was determined to be 0.814). CONCLUSIONS Plasma DNA is elevated in patients with deep vein thrombosis and correlates with biomarkers of DVT. A strong correlation between circulating DNA and MPO suggests that neutrophils may be a source of plasma DNA in patients with DVT.

The Role of Acidic Fibroblast Growth Factor and Fibroblast Growth Factor Receptor 4 in High Grade Serous Carcinoma

Background: High grade serous carcinoma whether ovarian, tubal or primary peritoneal, continues to be the most lethal gynecologic malignancy in the USA. Although combination chemotherapy and aggressive surgical resection has improved survival in the past decade the majority of patients still succumb to chemo-resistant disease recurrence. It has recently been reported that amplification of 5q31-5q35.3 is associated with poor prognosis in patients with high grade serous ovarian carcinoma. Although the amplicon contains over 50 genes, it is notable for the presence of several members of the fibroblast growth factor signaling axis. In particular acidic fibroblast growth factor (FGF1) has been demonstrated to be one of the driving genes in mediating the observed prognostic effect of the amplicon in ovarian cancer patients. This study seeks to further validate the prognostic value of fibroblast growth receptor 4 (FGFR4), another candidate gene of the FGF/FGFR axis located in the same amplicon. The emphasis will be delineating the role the FGF1/FGFR4 signaling axis plays in high grade serous ovarian carcinoma; and test the feasibility of targeting the FGF1/FGFR4 axis therapeutically. Materials and Methods: Spearman and Pearson correlation studies on data generated from array CGH and transcriptome profiling analyses on 51 microdissected tumor samples were used to identify genes located on chromosome 5q31-35.3 that showed significant correlation between DNA and mRNA copy numbers. Significant correlation between FGF1 and FGFR4 DNA copy numbers was further validated by qPCR analysis on DNA isolated from 51 microdissected tumor samples. Immunolocalization and quantification of FGFR4 expression were performed on paraffin embedded tissue samples from 183 cases of high-grade serous ovarian carcinoma. The expression was then correlated with clinical data to assess impact on survival. The expression of FGF1 and FGFR4 in vitro was quantified by real-time PCR and western blotting in six high-grade serous ovarian carcinoma cell lines and compared to those in human ovarian surface epithelial cells to identify overexpression. The effect of FGF1 on these cell lines after serum starvation was quantified for in vitro cellular proliferation, migration/invasion, chemoresistance and survival utilizing a combination of commercially available colorimetric, fluorometric and electrical impedance assays. FGFR4 expression was then transiently silenced via siRNA transfection and the effects on response to FGF1, cellular proliferation, and migration were quantified. To identify relevant cellular pathways involved, responsive cell lines were transduced with different transcription response elements using the Cignal-Lenti reporter system and treated with FGF1 with and without transient FGFR4 knock down. This was followed by western blot confirmation for the relevant phosphoproteins. Anti-FGF1 antibodies and FGFR trap proteins were used to attempt inhibition of FGF mediated phenotypic changes and relevant signaling in vitro. Orthotopic intraperitoneal tumors were established in nude mice using serous cell lines that have been previously transfected with luciferase expressing constructs. The mice were then treated with FGFR trap protein. Tumor progression was then followed via bioluminescent imaging. The FGFR4 gene from 52 clinical samples was sequenced to screen for mutations. Results: FGFR4 DNA and mRNA copy numbers were significantly correlated and FGFR4 DNA copy number was significantly correlated with that of FGF1. Survival of patients with high FGFR4 expressing tumors was significantly shorter that those with low expression(median survival 28 vs 55 month p< 0.001) In a multivariate cox regression model FGFR expression significantly increased risk of death (HR 2.1, p<0.001). FGFR4 expression was significantly higher in all cell lines tested compared to HOSE, OVCA432 cell line in particular had very high expression suggesting amplification. FGF1 was also particularly overexpressed in OVCA432. FGF1 significantly increased cell survival after serum deprivation in all cell lines. Transient knock down of FGFR4 caused significant reduction in cell migration and proliferation in vitro and significantly decreased the proliferative effects of FGF1 in vitro. FGFR1, FGFR4 traps and anti-FGF1 antibodies did not show activity in vitro. OVCA432 transfected with the cignal lenti reporter system revealed significant activation of MAPK, NFkB and WNT pathways, western blotting confirmed the results. Reverse phase protein array (RPPA) analysis also showed activation of MAPK, AKT, WNT pathways and down regulation of E Cadherin. FGFR trap protein significantly reduced tumor growth in vivo in an orthotopic mouse model. Conclusions: Overexpression and amplification of several members of the FGF signaling axis present on the amplicon 5q31-35.3 is a negative prognostic indicator in high grade serous ovarian carcinoma and may drive poor survival associated with that amplicon. Activation of The FGF signaling pathway leads to downstream activation of MAPK, AKT, WNT and NFkB pathways leading to a more aggressive cancer phenotype with increased tumor growth, evasion of apoptosis and increased migration and invasion. Inhibition of FGF pathway in vivo via FGFR trap protein leads to significantly decreased tumor growth in an orthotopic mouse model.

THE RELATIONSHIP BETWEEN OBESITY AND PSYCHOLOGICAL GENERAL WELL-BEING IN UNITED STATES WOMEN

While clinical studies have shown a negative relationship between obesity and mental health in women, population studies have not shown a consistent association. However, many of these studies can be criticized regarding fatness level criteria, lack of control variables, and validity of the psychological variables.^ The purpose of this research was to elucidate the relationship between fatness level and mental health in United States women using data from the First National Health and Nutrition Examination Survey (NHANES I), which was conducted on a national probability sample from 1971 to 1974. Mental health was measured by the General Well-Being Schedule (GWB), and fatness level was determined by the sum of the triceps and subscapular skinfolds. Women were categorized as lean (15th percentile or less), normal (16th to 84th percentiles), or obese (85th percentile or greater).^ A conceptual framework was developed which identified the variables of age, race, marital status, socioeconomic status (education), employment status, number of births, physical health, weight history, and perception of body image as important to the fatness level-GWB relationship. Multiple regression analyses were performed separately for whites and blacks with GWB as the response variable, and fatness level, age, education, employment status, number of births, marital status, and health perception as predictor variables. In addition, 2- and 3-way interaction terms for leanness, obesity and age were included as predictor variables. Variables related to weight history and perception of body image were not collected in NHANES I, and thus were not included in this study.^ The results indicated that obesity was a statistically significant predictor of lower GWB in white women even when the other predictor variables were controlled. The full regression model identified the young, more educated, obese female as a subgroup with lower GWB, especially in blacks. These findings were not consistent with the previous non-clinical studies which found that obesity was associated with better mental health. The social stigma of being obese and the preoccupation of women with being lean may have contributed to the lower GWB in these women. ^

Biodiversity effects on nitrate concentrations in soil solution: a Bayesian model

Ecosystems are faced with high rates of species loss which has consequences for their functions and services. To assess the effects of plant species diversity on the nitrogen (N) cycle, we developed a model for monthly mean nitrate (NO3-N) concentrations in soil solution in 0-30 cm mineral soil depth using plant species and functional group richness and functional composition as drivers and assessing the effects of conversion of arable land to grassland, spatially heterogeneous soil properties, and climate. We used monthly mean NO3-N concentrations from 62 plots of a grassland plant diversity experiment from 2003 to 2006. Plant species richness (1-60) and functional group composition (1-4 functional groups: legumes, grasses, non-leguminous tall herbs, non-leguminous small herbs) were manipulated in a factorial design. Plant community composition, time since conversion from arable land to grassland, soil texture, and climate data (precipitation, soil moisture, air and soil temperature) were used to develop one general Bayesian multiple regression model for the 62 plots to allow an in-depth evaluation using the experimental design. The model simulated NO3-N concentrations with an overall Bayesian coefficient of determination of 0.48. The temporal course of NO3-N concentrations was simulated differently well for the individual plots with a maximum plot-specific Nash-Sutcliffe Efficiency of 0.57. The model shows that NO3-N concentrations decrease with species richness, but this relation reverses if more than approx. 25 % of legume species are included in the mixture. Presence of legumes increases and presence of grasses decreases NO3-N concentrations compared to mixtures containing only small and tall herbs. Altogether, our model shows that there is a strong influence of plant community composition on NO3-N concentrations.