Semi-supervised dimensionality reduction based on partial least squares for visual analysis of high dimensional data

Dimensionality reduction is employed for visual data analysis as a way to obtaining reduced spaces for high dimensional data or to mapping data directly into 2D or 3D spaces. Although techniques have evolved to improve data segregation on reduced or visual spaces, they have limited capabilities for adjusting the results according to user's knowledge. In this paper, we propose a novel approach to handling both dimensionality reduction and visualization of high dimensional data, taking into account user's input. It employs Partial Least Squares (PLS), a statistical tool to perform retrieval of latent spaces focusing on the discriminability of the data. The method employs a training set for building a highly precise model that can then be applied to a much larger data set very effectively. The reduced data set can be exhibited using various existing visualization techniques. The training data is important to code user's knowledge into the loop. However, this work also devises a strategy for calculating PLS reduced spaces when no training data is available. The approach produces increasingly precise visual mappings as the user feeds back his or her knowledge and is capable of working with small and unbalanced training sets.

A general class of zero-or-one inflated beta regression models

This paper proposes a general class of regression models for continuous proportions when the data contain zeros or ones. The proposed class of models assumes that the response variable has a mixed continuous-discrete distribution with probability mass at zero or one. The beta distribution is used to describe the continuous component of the model, since its density has a wide range of different shapes depending on the values of the two parameters that index the distribution. We use a suitable parameterization of the beta law in terms of its mean and a precision parameter. The parameters of the mixture distribution are modeled as functions of regression parameters. We provide inference, diagnostic, and model selection tools for this class of models. A practical application that employs real data is presented. (C) 2011 Elsevier B.V. All rights reserved.

Random regression analyses using B-spline functions to model growth of Nellore cattle

The objective of this study was to estimate (co)variance components using random regression on B-spline functions to weight records obtained from birth to adulthood. A total of 82 064 weight records of 8145 females obtained from the data bank of the Nellore Breeding Program (PMGRN/Nellore Brazil) which started in 1987, were used. The models included direct additive and maternal genetic effects and animal and maternal permanent environmental effects as random. Contemporary group and dam age at calving (linear and quadratic effect) were included as fixed effects, and orthogonal Legendre polynomials of age (cubic regression) were considered as random covariate. The random effects were modeled using B-spline functions considering linear, quadratic and cubic polynomials for each individual segment. Residual variances were grouped in five age classes. Direct additive genetic and animal permanent environmental effects were modeled using up to seven knots (six segments). A single segment with two knots at the end points of the curve was used for the estimation of maternal genetic and maternal permanent environmental effects. A total of 15 models were studied, with the number of parameters ranging from 17 to 81. The models that used B-splines were compared with multi-trait analyses with nine weight traits and to a random regression model that used orthogonal Legendre polynomials. A model fitting quadratic B-splines, with four knots or three segments for direct additive genetic effect and animal permanent environmental effect and two knots for maternal additive genetic effect and maternal permanent environmental effect, was the most appropriate and parsimonious model to describe the covariance structure of the data. Selection for higher weight, such as at young ages, should be performed taking into account an increase in mature cow weight. Particularly, this is important in most of Nellore beef cattle production systems, where the cow herd is maintained on range conditions. There is limited modification of the growth curve of Nellore cattle with respect to the aim of selecting them for rapid growth at young ages while maintaining constant adult weight.

Limited value of ultrasound assessment in patients with poor outcome after carpal tunnel release surgerys

[EN] OBJECTIVE: To determine the value of ultrasonography in the assessment of patients with idiopathic carpal tunnel syndrome (CTS) and poor outcome after carpal tunnel release. METHODS: A total of 88 consecutive patients with CTS (104 hands) underwent open surgical release of the median nerve. Ultrasound (US) examination was performed blind to any patient's data. The median nerve area at tunnel inlet and outlet, the retinaculum distance, and the flattening ratio were measured. The main outcome variable was the patient's overall satisfaction using a five-point Likert scale (1 = worse, 2 = no change, 3 = slightly better, 4 = much better, 5 = cured) at 3 months postoperatively. Pre- and postoperative ultrasonographic findings in relation to clinical outcome were analysed. RESULTS: Improvement (scores 4 or 5 on the Likert scale) was recorded in 75 hands (72%). After carpal tunnel release, the cross-sectional area at tunnel inlet decreased from a mean of 14.2 to 13.3 mm2 in the group with clinical improvement and also from a mean of 12.5 to 11.6 mm2 in the group with no change or slight improvement. No significant changes in the cross-sectional area at tunnel outlet, retinaculum distance, and flattening ratio were observed. CONCLUSION: Reduction of the median nerve cross-sectional area at tunnel inlet at 3 months after carpal tunnel release was similar in patients reporting cure or great improvement and in those with slight or no improvement. Ultrasonography is of limited value in assessment of patients with poor outcome after median nerve release.

Dose-dependent effect of a single GnRHa injection on the spawning of meagre ("Argyrosomus regius") broodstock reared in captivity

[EN]The present study aimed to determine the spawning efficacy, egg quality and quantity of captive breed meagre induced with a single gonadotrophin-releasing hormone agonist (GnRHa) injection of 0, 1, 5, 10, 15, 20, 25, 30, 40 or 50 μg kg–1 to determine a recommended optimum dose to induce spawning. The doses 10, 15 and 20 μg kg–1 gave eggs with the highest quality (measured as: percentage of viability, floating, fertilisation and hatch) and quantity (measured as: total number of eggs, number of viable eggs, number of floating eggs, number of hatched larvae and number of larvae that reabsorbed the yolk sac). All egg quantity parameters were described by Gaussian regression analysis with R2 = 0.89 or R2 = 0.88. The Gaussian regression analysis identified that the optimal dose used was 15 μg kg–1. The regression analysis highlighted that this comprehensive study examined doses that ranged from low doses insufficient to stimulate a high spawning response (significantly lower egg quantities, p < 0.05) compared to 15 μg kg–1 through to high doses that stimulated the spawning of significantly lower egg quantities and eggs with significantly lower quality (egg viability). In addition, the latency period (time from hormone application to spawning) decreased with increasing doses to give a regression (R2 = 0.93), which suggests that higher doses accelerated oocyte development that in turn reduced egg quality and quantity. The identification of an optimal dose for the spawning of meagre, which has high aquaculture potential, represents an important advance for the Mediterranean aquaculture industry.

NMR, Metabonomics and Molecular Profiles: Applications to the Quality Assessment of Foodstuffs

Nuclear Magnetic Resonance (NMR) is a branch of spectroscopy that is based on the fact that many atomic nuclei may be oriented by a strong magnetic field and will absorb radiofrequency radiation at characteristic frequencies. The parameters that can be measured on the resulting spectral lines (line positions, intensities, line widths, multiplicities and transients in time-dependent experi-ments) can be interpreted in terms of molecular structure, conformation, molecular motion and other rate processes. In this way, high resolution (HR) NMR allows performing qualitative and quantitative analysis of samples in solution, in order to determine the structure of molecules in solution and not only. In the past, high-field NMR spectroscopy has mainly concerned with the elucidation of chemical structure in solution, but today is emerging as a powerful exploratory tool for probing biochemical and physical processes. It represents a versatile tool for the analysis of foods. In literature many NMR studies have been reported on different type of food such as wine, olive oil, coffee, fruit juices, milk, meat, egg, starch granules, flour, etc using different NMR techniques. Traditionally, univariate analytical methods have been used to ex-plore spectroscopic data. This method is useful to measure or to se-lect a single descriptive variable from the whole spectrum and , at the end, only this variable is analyzed. This univariate methods ap-proach, applied to HR-NMR data, lead to different problems due especially to the complexity of an NMR spectrum. In fact, the lat-ter is composed of different signals belonging to different mole-cules, but it is also true that the same molecules can be represented by different signals, generally strongly correlated. The univariate methods, in this case, takes in account only one or a few variables, causing a loss of information. Thus, when dealing with complex samples like foodstuff, univariate analysis of spectra data results not enough powerful. Spectra need to be considered in their wholeness and, for analysing them, it must be taken in consideration the whole data matrix: chemometric methods are designed to treat such multivariate data. Multivariate data analysis is used for a number of distinct, differ-ent purposes and the aims can be divided into three main groups: • data description (explorative data structure modelling of any ge-neric n-dimensional data matrix, PCA for example); • regression and prediction (PLS); • classification and prediction of class belongings for new samples (LDA and PLS-DA and ECVA). The aim of this PhD thesis was to verify the possibility of identify-ing and classifying plants or foodstuffs, in different classes, based on the concerted variation in metabolite levels, detected by NMR spectra and using the multivariate data analysis as a tool to inter-pret NMR information. It is important to underline that the results obtained are useful to point out the metabolic consequences of a specific modification on foodstuffs, avoiding the use of a targeted analysis for the different metabolites. The data analysis is performed by applying chemomet-ric multivariate techniques to the NMR dataset of spectra acquired. The research work presented in this thesis is the result of a three years PhD study. This thesis reports the main results obtained from these two main activities: A1) Evaluation of a data pre-processing system in order to mini-mize unwanted sources of variations, due to different instrumental set up, manual spectra processing and to sample preparations arte-facts; A2) Application of multivariate chemiometric models in data analy-sis.

Spinal cord injury: assessment of autonomic state-dependent control of cardiovascular system and body core temperature

Spinal cord injury (SCI) results not only in paralysis; but it is also associated with a range of autonomic dysregulation that can interfere with cardiovascular, bladder, bowel, temperature, and sexual function. The entity of the autonomic dysfunction is related to the level and severity of injury to descending autonomic (sympathetic) pathways. For many years there was limited awareness of these issues and the attention given to them by the scientific and medical community was scarce. Yet, even if a new system to document the impact of SCI on autonomic function has recently been proposed, the current standard of assessment of SCI (American Spinal Injury Association (ASIA) examination) evaluates motor and sensory pathways, but not severity of injury to autonomic pathways. Beside the severe impact on quality of life, autonomic dysfunction in persons with SCI is associated with increased risk of cardiovascular disease and mortality. Therefore, obtaining information regarding autonomic function in persons with SCI is pivotal and clinical examinations and laboratory evaluations to detect the presence of autonomic dysfunction and quantitate its severity are mandatory. Furthermore, previous studies demonstrated that there is an intimate relationship between the autonomic nervous system and sleep from anatomical, physiological, and neurochemical points of view. Although, even if previous epidemiological studies demonstrated that sleep problems are common in spinal cord injury (SCI), so far only limited polysomnographic (PSG) data are available. Finally, until now, circadian and state dependent autonomic regulation of blood pressure (BP), heart rate (HR) and body core temperature (BcT) were never assessed in SCI patients. Aim of the current study was to establish the association between the autonomic control of the cardiovascular function and thermoregulation, sleep parameters and increased cardiovascular risk in SCI patients.

In vivo and in vitro investigation of the tissue-specific activity of the human CYP3A4 and CYP3A5 promoters

The human cytochrome P450 3A4 (CYP3A4), the predominant but variably expressed cytochrome P450 in adult liver and small intestine is involved in the metabolism of over 50% of currently used drugs. Its paralog CYP3A5 plays a crucial role in the disposition of several drugs with low therapeutic index, including tacrolimus. Limited information is available for the CYP3A5 transcriptional regulation and its induction by xenobiotics remains controversial. In the first part of this study, we analysed the CYP3A5 transcriptional regulation and its induction by xenobiotics in vivo using transgenic mice. To this end, two transgenic strains were established by pronuclear injection of a plasmid, expressing firefly luciferase driven by a 6.2 kb of the human CYP3A5 promoter. A detailed analysis of both strains shows a tissue distribution largely reflecting that of CYP3A5 transcripts in humans. Thus, the highest luciferase activity was detected in the small intestine, followed by oesophagus, testis, lung, adrenal gland, ovary, prostate and kidney. However, no activity was observed in the liver. CYP3A5-luc transgenic mice were similarly induced in both sexes with either PCN or TCPOBOP in small intestine in a dose-dependent manner. Thus, the 6.2 kb upstream promoter of CYP3A5 mediates the broad tissue activity in transgenic mice. CYP3A5 promoter is inducible in the small intestine in vivo, which may contribute to the variable expression of CYP3A in this organ. rnThe hepato-intestinal level of the detoxifying oxidases CYP3A4 and CYP3A5 is adjusted to the xenobiotic exposure mainly via the xenosensor and transcriptional factor PXR. CYP3A5 is additionally expressed in several other organs lacking PXR, including kidney. In the second part of this study, we investigated the mechanism of the differential expression of CYP3A5 and CYP3A4 and its evolutionary origin using renal and intestinal cells, and comparative genomics. For this examination, we established a two-cell line models reflecting the expression relationships of CYP3A4 and CYP3A5 in the kidney and small intestine in vivo. Our data demonstrate that the CYP3A5 expression in renal cells was enabled by the loss of a suppressing Yin Yang 1 (YY1)-binding site from the CYP3A5 promoter. This allowed for a renal CYP3A5 expression in a PXR-independent manner. The YY1 element is retained in the CYP3A4 gene, leading to its suppression, perhaps via interference with the NF1 activity in renal cells. In intestinal cells, the inhibition of CYP3A4 expression by YY1 is abrogated by a combined activating effect of PXR and NF1 acting on their respective response elements located adjacent to the YY1-binding site on CYP3A4 proximal promoter. CYP3A4 expression is further facilitated by a point mutation attenuating the suppressing effect of YY1 binding site. The differential expression of CYP3A4 and CYP3A5 in these organs results from the loss of the YY1 binding element from the CYP3A5 promoter, acting in concert with the differential organ expression of PXR, and with the higher accumulation of PXR response elements in CYP3A4. rn

Bio-inspired polymer-supported nitrido molybdenum(VI) complexes for ammonia synthesis − reactivity towards protons and electrons : EPR investigations on paramagnetic molybdenum(V) and copper(II) complexes

Die biologische Stickstofffixierung durch Molybdän-haltige Nitrogenasen sowie die Erforschung des zugrundeliegenden komplexen Mechanismus (N2-Aktivierung an Metall-Zentren, 6-fache Protonierung und Reduktion, N–N Bindungsspaltung unter Bildung von Ammoniak) ist von erheblichem Interesse. Insbesondere Molybdän-Komplexe wurden bereits erfolgreich als Modellverbindungen für die Untersuchung elementarer Einzelschritte der N2-Aktivierung eingesetzt. Durch die Verwendung von Triamidoamin-Liganden ist es Schrock et al. sogar gelungen mehrere Katalysezyklen zu durchlaufen und einen Mechanismus zu formulieren. Trotz der sterisch anspruchsvollen Substituenten in den Schrock-Komplexen ist die Umsatzrate dieses homogenen Katalysators, aufgrund Komplex-Deaktivierung infolge intermolekularer Reaktionen wie Dimerisierung und Disproportionierung, limitiert. In der vorliegenden Arbeit wurden einige dieser Herausforderungen angegangen und die aktiven Spezies auf einer Festphase immobilisiert, um intermolekulare Reaktionen durch räumliche Isolierung der Komplexe zu unterdrücken.rnEin Polymer-verankertes Analogon des Schrock Nitrido-Molybdän(VI)-Komplexes wurde auf einem neuen Reaktionsweg synthetisiert. Dieser beinhaltet nur einen einzigen Reaktionsschritt, um die funktionelle Gruppe „MoN“ einzuführen. Protonierung des immobilisierten Nitrido-Molybdän(VI)-Komplexes LMoVIN (L = Polymer-verankerter Triamidoamin-Ligand) mit 2,6-Lutidinium liefert den entsprechenden Imido-Molybdän(VI)-Komplex. Durch anschließende Ein-Elektronen-Reduktion mit Cobaltocen wird der Polymer-angebundene Imido-Molybdän(V)-Komplex erhalten, bewiesen durch EPR-Spektroskopie (g1,2,3 = 1.989, 1.929, 1.902). Durch die Immobilisierung und die effektive räumliche Separation der Reaktionszentren auf der Festphase werden bimolekulare Nebenreaktionen, die oft in homogenen Systemen auftreten, unterdrückt. Dies ermöglicht zum ersten Mal die Darstellung des Imido-Molybdän(V)-Intermediates des Schrock-Zyklus.rnEPR-Spektren des als Spin-Label eingeführten immobilisierten Nitrato-Kupfer(II)-Komplexes wurden unter verschiedenen Bedingungen (Lösungsmittel, Temperatur) aufgenommen, wobei sich eine starke Abhängigkeit zwischen der Zugänglichkeit und Reaktivität der immobilisierten Reaktionszentren und der Art des Lösungsmittels zeigte. Somit wurde die Reaktivität von LMoVIN gegenüber Protonen und Elektronen, welches zur Bildung von NH3 führt, unter Verwendung verschiedener Lösungsmittel untersucht und optimiert. Innerhalb des kugelförmigen Polymers verläuft die Protonierung und Reduktion von LMoVIN stufenweise. Aktive Zentren, die sich an der „äußeren Schale“ des Polymers befinden, sind gut zugänglich und reagieren schnell nach H+/e− Zugabe. Aktive Zentren im „Inneren des Polymers“ hingegen sind schlechter zugänglich und zeigen langsame diffusions-kontrollierte Reaktionen, wobei drei H+/e− Schritte gefolgt von einer Ligandenaustausch-Reaktion erforderlich sind, um NH3 freizusetzen: LMoVIN  LMoVNH  LMoIVNH2  LMoIIINH3 und anschließender Ligandenaustausch führt zur Freisetzung von NH3.rnIn einem weiteren Projekt wurde der Bis(ddpd)-Kupfer(II)-Komplex EPR-spektroskopisch in Hinblick auf Jahn−Teller-Verzerrung und -Dynamik untersucht. Dabei wurden die EPR-Spektren bei variabler Temperatur (70−293 K) aufgenommen. Im Festkörperspektrum bei T &lt; 100 K erscheint der Kupfer(II)-Komplex als gestreckter Oktaeder, wohingegen das EPR-Spektrum bei höheren Temperaturen g-Werte aufzeigt, die einer pseudo-gestauchten oktaedrischen Kupfer(II)-Spezies zuzuordnen sind. Diese Tatsache wird einem intramolekularen dynamischen Jahn−Teller Phänomen zugeschrieben, welcher bei 100 K eingefroren wird.

Pericellular fibronectin is required for RhoA-dependent responses to cyclic strain in fibroblasts

To test the hypothesis that the pericellular fibronectin matrix is involved in mechanotransduction, we compared the response of normal and fibronectin-deficient mouse fibroblasts to cyclic substrate strain. Normal fibroblasts seeded on vitronectin in fibronectin-depleted medium deposited their own fibronectin matrix. In cultures exposed to cyclic strain, RhoA was activated, actin-stress fibers became more prominent, MAL/MKL1 shuttled to the nucleus, and mRNA encoding tenascin-C was induced. By contrast, these RhoA-dependent responses to cyclic strain were suppressed in fibronectin knockdown or knockout fibroblasts grown under identical conditions. On vitronectin substrate, fibronectin-deficient cells lacked fibrillar adhesions containing alpha5 integrin. However, when fibronectin-deficient fibroblasts were plated on exogenous fibronectin, their defects in adhesions and mechanotransduction were restored. Studies with fragments indicated that both the RGD-synergy site and the adjacent heparin-binding region of fibronectin were required for full activity in mechanotransduction, but not its ability to self-assemble. In contrast to RhoA-mediated responses, activation of Erk1/2 and PKB/Akt by cyclic strain was not affected in fibronectin-deficient cells. Our results indicate that pericellular fibronectin secreted by normal fibroblasts is a necessary component of the strain-sensing machinery. Supporting this hypothesis, induction of cellular tenascin-C by cyclic strain was suppressed by addition of exogenous tenascin-C, which interferes with fibronectin-mediated cell spreading.

Co2-dependent vasomotor reactivity of cerebral arteries in patients with severe traumatic brain injury: time course and effect of augmentation of cardiac output with dobutamine

Failing cerebral blood flow (CBF) autoregulation may contribute to cerebral damage after traumatic brain injury (TBI). The purpose of this study was to describe the time course of CO(2)-dependent vasoreactivity, measured as CBF velocity in response to hyperventilation (vasomotor reactivity [VMR] index). We included 13 patients who had had severe TBI, 8 of whom received norepinephrine (NE) based on clinical indication. In these patients, measurements were also performed after dobutamine administration, with a goal of increasing cardiac output by 30%. Blood flow velocity was measured with transcranial Doppler ultrasound in both hemispheres. All patients except one had an abnormal VMR index in at least one hemisphere within the first 24 h after TBI. In those patients who did not receive catecholamines, mean VMR index recovered within the first 48 to 72 h. In contrast, in patients who received NE within the first 48 h period, VMR index did not recover on the second day. Cardiac output and mean CBF velocity increased significantly during dobutamine administration, but VMR index did not change significantly. In conclusion, CO(2) vasomotor reactivity was abnormal in the first 24 h after TBI in most of the patients, but recovered within 48 h in those patients who did not receive NE, in contrast to those eventually receiving the drug. Addition of dobutamine to NE had variable but overall insignificant effects on CO(2) vasomotor reactivity.

Experimental Evaluation of the Influence of Human-Structure Interaction for Vibration Serviceability

The effects of human-structure interaction on the dynamic performance of occupied structures have long been observed. The inclusion of the effects of human-structure interaction is important to ensure that the dynamic response of a structure is not overestimated. Previous observations, both in service and in the laboratory, have yielded results indicating that the effects are dependent on the natural frequency of the structure, the posture of the occupants, and the mass ratio of the occupants to the structure. These results are noteworthy, but are limited in their application,because the data are sparse and are only pertinent to a specific set of characteristics identified in a given study. To examine these characteristics simultaneously and consistently, an experimental test structure was designed with variable properties to replicate a variety of configurations within a controlled setting focusing on the effects of passive occupants. Experimental modal analysis techniques were employed to both the empty and occupied conditions of the structure and the dynamic properties associated with each condition were compared. Results similar to previous investigations were observed, including both an increase and a decrease in natural frequency of the occupied structure with respect to the empty structure, as well as the identification of a second mode of vibration. The damping of the combined system was higher for all configurations. Overall, this study provides a broad data set representing a wide array of configurations. The experimental results of this study were used to assess current recommendations for the dynamic properties of a crowd to analytically predict the effects of human-structure interaction. The experimental results were used to select a set of properties for passive, standing occupants and develop a new model that can more accurately represent the behavior of the human-structure system as experimentally measured in this study.

Prognostic impact of lymph node ratio outperforms positive lymph nodes and lymph nodes harvested: a time-dependent analysis in mismatch repair-proficient and –deficient colrectal cancers

Objective: We compare the prognostic strength of the lymph node ratio (LNR), positive lymph nodes (+LNs) and collected lymph nodes (LNcoll) using a time-dependent analysis in colorectal cancer patients stratified by mismatch repair (MMR) status. Method: 580 stage III-IV patients were included. Multivariable Cox regression analysis and time-dependent receiver operating characteristic (tROC) curve analysis were performed. The Area under the Curve (AUC) over time was compared for the three features. Results were validated on a second cohort of 105 stage III-IV patients. Results: The AUC for the LNR was 0.71 and outperformed + LNs and LNcoll by 10–15 % in both MMR-proficient and deficient cancers. LNR and + LNs were both significant (p<0.0001) in multivariable analysis but the effect was considerably stronger for the LNR [LNR: HR=5.18 (95 % CI: 3.5–7.6); +LNs=1.06 (95 % CI: 1.04–1.08)]. Similar results were obtained for patients with >12 LNcoll. An optimal cut off score for LNR=0.231 was validated on the second cohort (p<0.001). Conclusion: The LNR outperforms the + LNs and LNcoll even in patients with >12 LNcoll. Its clinical value is not confounded by MMR status. A cut-of score of 0.231 may best stratify patients into prognostic subgroups and could be a basis for the future prospective analysis of the LNR.

Leydig-cell tumour in children: variable clinical presentation, diagnostic features, follow-up and genetic analysis of four cases

BACKGROUND: Testicular tumours are relatively uncommon in infants and children, accounting for only 1-2% of all paediatric solid tumours. Of these approximately 1.5% are Leydig-cell tumours. Further, activating mutations of the luteinizing hormone receptor gene (LHR), as well as of the G protein genes, such as Gsalpha (gsp) and Gialpha (gip2) subunits, and cyclin-dependent kinase gene 4(CDK4) have been associated with the development of several endocrine neoplasms. AIMS/METHODS: In this report, the clinical variability of Leydig-cell tumours in four children is described. The LHR-, gsp-, gip2- and CDK4 genes were investigated to establish the possible molecular pathogenesis of the variable phenotype of the Leydig-cell tumours. RESULTS: No activating mutations in these genes were found in the four Leydig-cell tumours studied. Therefore, the absence of activating mutations in LHR, as well as in both the 'hot spot' regions for activating mutations within the G-alpha subunits and in the regulatory 'hot spot' on the CDK4 genes in these tumours indicates molecular heterogeneity among Leydig-cell tumours. CONCLUSION: Four children with a variable phenotype caused by Leydig-cell tumours are described. A molecular analysis of all the 'activating' genes and mutational regions known so far was performed, but no abnormalities were found. The lessons learnt from these clinically variable cases are: perform ultrasound early and most importantly, consider discrepancies between testicular swelling, tumour size and androgen production.

Locally Efficient Estimation in Censored Data Models: Theory and Examples

In many applications the observed data can be viewed as a censored high dimensional full data random variable X. By the curve of dimensionality it is typically not possible to construct estimators that are asymptotically efficient at every probability distribution in a semiparametric censored data model of such a high dimensional censored data structure. We provide a general method for construction of one-step estimators that are efficient at a chosen submodel of the full-data model, are still well behaved off this submodel and can be chosen to always improve on a given initial estimator. These one-step estimators rely on good estimators of the censoring mechanism and thus will require a parametric or semiparametric model for the censoring mechanism. We present a general theorem that provides a template for proving the desired asymptotic results. We illustrate the general one-step estimation methods by constructing locally efficient one-step estimators of marginal distributions and regression parameters with right-censored data, current status data and bivariate right-censored data, in all models allowing the presence of time-dependent covariates. The conditions of the asymptotics theorem are rigorously verified in one of the examples and the key condition of the general theorem is verified for all examples.