Differential Dimerization of Variants Linked to Enhanced S-Cone Sensitivity Syndrome (ESCS) Located in the NR2E3 Ligand-Binding Domain.

NR2E3 encodes the photoreceptor-specific nuclear hormone receptor that acts as a repressor of cone-specific gene expression in rod photoreceptors, and as an activator of several rod-specific genes. Recessive variants located in the ligand-binding domain (LBD) of NR2E3 cause enhanced short wavelength sensitive- (S-) cone syndrome (ESCS), a retinal degeneration characterized by an excess of S-cones and non-functional rods. We analyzed the dimerization properties of NR2E3 and the effect of disease-causing LBD missense variants by bioluminescence resonance energy transfer (BRET(2) ) protein interaction assays. Homodimerization was not affected in presence of p.A256V, p.R039G, p.R311Q, and p.R334G variants, but abolished in presence of p.L263P, p.L336P, p.L353V, p.R385P, and p.M407K variants. Homology modeling predicted structural changes induced by NR2E3 LBD variants. NR2E3 LBD variants did not affect interaction with CRX, but with NRL and rev-erbα/NR1D1. CRX and NRL heterodimerized more efficiently together, than did either with NR2E3. NR2E3 did not heterodimerize with TLX/NR2E1 and RXRα/NR2C1. The identification of a new compound heterozygous patient with detectable rod function, who expressed solely the p.A256V variant protein, suggests a correlation between LBD variants able to form functional NR2E3 dimers and atypical mild forms of ESCS with residual rod function.

Pointwise estimates for the Bergman kernel of the weighted Fock space

We prove upper pointwise estimates for the Bergman kernel of the weighted Fock space of entire functions in $L^{2}(e^{-2\phi}) $ where $\phi$ is a subharmonic function with $\Delta\phi$ a doubling measure. We derive estimates for the canonical solution operator to the inhomogeneous Cauchy-Riemann equation and we characterize the compactness of this operator in terms of $\Delta\phi$.

Kernel-PCA data integration with enhanced interpretability

Background Nowadays, combining the different sources of information to improve the biological knowledge available is a challenge in bioinformatics. One of the most powerful methods for integrating heterogeneous data types are kernel-based methods. Kernel-based data integration approaches consist of two basic steps: firstly the right kernel is chosen for each data set; secondly the kernels from the different data sources are combined to give a complete representation of the available data for a given statistical task. Results We analyze the integration of data from several sources of information using kernel PCA, from the point of view of reducing dimensionality. Moreover, we improve the interpretability of kernel PCA by adding to the plot the representation of the input variables that belong to any dataset. In particular, for each input variable or linear combination of input variables, we can represent the direction of maximum growth locally, which allows us to identify those samples with higher/lower values of the variables analyzed. Conclusions The integration of different datasets and the simultaneous representation of samples and variables together give us a better understanding of biological knowledge.

Neutralization of (NK-cell-derived) B-cell activating factor by Belimumab restores sensitivity of chronic lymphoid leukemia cells to direct and Rituximab-induced NK lysis.

Natural killer (NK) cells are cytotoxic lymphocytes that substantially contribute to the therapeutic benefit of antitumor antibodies like Rituximab, a crucial component in the treatment of B-cell malignancies. In chronic lymphocytic leukemia (CLL), the ability of NK cells to lyse the malignant cells and to mediate antibody-dependent cellular cytotoxicity upon Fc receptor stimulation is compromised, but the underlying mechanisms are largely unclear. We report here that NK-cells activation-dependently produce the tumor necrosis factor family member 'B-cell activating factor' (BAFF) in soluble form with no detectable surface expression, also in response to Fc receptor triggering by therapeutic CD20-antibodies. BAFF in turn enhanced the metabolic activity of primary CLL cells and impaired direct and Rituximab-induced lysis of CLL cells without affecting NK reactivity per se. The neutralizing BAFF antibody Belimumab, which is approved for treatment of systemic lupus erythematosus, prevented the effects of BAFF on the metabolism of CLL cells and restored their susceptibility to direct and Rituximab-induced NK-cell killing in allogeneic and autologous experimental systems. Our findings unravel the involvement of BAFF in the resistance of CLL cells to NK-cell antitumor immunity and Rituximab treatment and point to a benefit of combinatory approaches employing BAFF-neutralizing drugs in B-cell malignancies.

Beurling-Landau densities of weighted Fekete sets and correlation kernel estimates

Let $Q$ be a suitable real function on $C$. An $n$-Fekete set corresponding to $Q$ is a subset ${Z_{n1}},\dotsb, Z_{nn}}$ of $C$ which maximizes the expression $\Pi^n_i_{

Tumour biology, metastatic sites and taxanes sensitivity as determinants of eribulin mesylate efficacy in breast cancer: results from the ERIBEX retrospective, international, multicenter study.

BACKGROUND: Our retrospective, international study aimed at evaluating the activity and safety of eribulin mesylate (EM) in pretreated metastatic breast cancer (MBC) in a routine clinical setting. METHODS: Patients treated with EM for a locally advanced or MBC between March 2011 and January 2014 were included in the study. Clinical and biological assessment of toxicity was performed at each visit. Tumour response was assessed every 3 cycles of treatment. A database was created to collect clinical, pathological and treatment data. RESULTS: Two hundred and fifty-eight patients were included in the study. Median age was 59 years old. Tumours were Hormone Receptor (HR)-positive (73.3 %) HER2-positive (10.2 %), and triple negative (TN, 22.5 %). 86.4 % of the patients presented with visceral metastases, mainly in the liver (67.4 %). Median previous metastatic chemotherapies number was 4 [1-9]. Previous treatments included anthracyclines and/or taxanes (100 %) and capecitabine (90.7 %). Median number of EM cycles was 5 [1-19]. The relative dose intensity was 0.917. At the time of analysis (median follow-up of 13.9 months), 42.3 % of the patients were still alive. The objective response rate was 25.2 % (95 %CI: 20-31) with a 36.1 % clinical benefit rate (CBR). Median time to progression (TTP) and overall survival were 3.97 (95 %CI: 3.25-4.3) and 11.2 (95 %CI: 9.3-12.1) months, respectively. One- and 2-year survival rates were 45.5 and 8.5 %, respectively. In multivariate analysis, HER2 positivity (HR = 0.29), the presence of lung metastases (HR = 2.49) and primary taxanes resistance (HR = 2.36) were the only three independent CBR predictive factors, while HR positivity (HR = 0.67), the presence of lung metastases (HR = 1.52) and primary taxanes resistance (HR = 1.50) were the only three TTP independent prognostic factors. Treatment was globally well tolerated. Most common grade 3-4 toxicities were neutropenia (20.9 %), peripheral neuropathy (3.9 %), anaemia (1.6 %), liver dysfunction (0.8 %) and thrombocytopenia (0.4 %). Thirteen patients (5 %) developed febrile neutropenia. CONCLUSION: EM is an effective new option in heavily pretreated MBC, with a favourable efficacy/safety ratio in a clinical practice setting. Our results comfort the use of this new molecule and pledge for the evaluation of EM-trastuzumab combination in this setting. Tumour biology, primary taxanes sensitivity and metastatic sites could represent useful predictive and prognostic factors.

New kernel functions and learning methods for text and data mining

Recent advances in machine learning methods enable increasingly the automatic construction of various types of computer assisted methods that have been difficult or laborious to program by human experts. The tasks for which this kind of tools are needed arise in many areas, here especially in the fields of bioinformatics and natural language processing. The machine learning methods may not work satisfactorily if they are not appropriately tailored to the task in question. However, their learning performance can often be improved by taking advantage of deeper insight of the application domain or the learning problem at hand. This thesis considers developing kernel-based learning algorithms incorporating this kind of prior knowledge of the task in question in an advantageous way. Moreover, computationally efficient algorithms for training the learning machines for specific tasks are presented. In the context of kernel-based learning methods, the incorporation of prior knowledge is often done by designing appropriate kernel functions. Another well-known way is to develop cost functions that fit to the task under consideration. For disambiguation tasks in natural language, we develop kernel functions that take account of the positional information and the mutual similarities of words. It is shown that the use of this information significantly improves the disambiguation performance of the learning machine. Further, we design a new cost function that is better suitable for the task of information retrieval and for more general ranking problems than the cost functions designed for regression and classification. We also consider other applications of the kernel-based learning algorithms such as text categorization, and pattern recognition in differential display. We develop computationally efficient algorithms for training the considered learning machines with the proposed kernel functions. We also design a fast cross-validation algorithm for regularized least-squares type of learning algorithm. Further, an efficient version of the regularized least-squares algorithm that can be used together with the new cost function for preference learning and ranking tasks is proposed. In summary, we demonstrate that the incorporation of prior knowledge is possible and beneficial, and novel advanced kernels and cost functions can be used in algorithms efficiently.

(MA)LDI MS Imaging at High Specificity and Sensitivity

(Matrix-assisted) laser desorption/ionization ((MA)LDI) mass spectrometry imaging (MSI) has been driven by remarkable technological developments in the last couple of years. Although molecular information of a wide range of molecules including peptides, lipids, metabolites, and xenobiotics can be mapped, (MA)LDI MSI only leads to the detection of the most abundant soluble molecules in the cells and, consequently, does not provide access to the least expressed species, which can be very informative in the scope of disease research. Within a short period of time, numerous protocols and concepts have been developed and introduced in order to increase MSI sensitivity, including in situ tissue chemistry and solvent-free matrix depositions. In this chapter, we will discuss some of the latest developments in the field of high-sensitivity MSI using solvent-free matrix depositions and will detail protocols of two methods with their capability of enriching molecular MSI signal as demonstrated within our laboratory.

Maladaptive emotion regulation strategies and stress sensitivity mediate the relation between adverse life events and attenuated positive psychotic symptoms.

INTRODUCTION: There is now solid evidence for a relation between adverse life events (ALE) and psychotic symptoms in patients with psychosis and in the general population. A recent study has shown that this relation may be partially mediated by stress sensitivity, suggesting the influence of other factors. The aim of this study was to assess the mediation effect of emotion regulation strategies and stress sensitivity in the relation between ALE and attenuated positive psychotic symptoms (APPS) in the general population. METHODS: Hundred and twelve healthy volunteers were evaluated with measures of APPS, emotion regulation strategies, ALE and stress sensitivity. RESULTS: Results demonstrated that the relation between ALE, hallucination and delusion proneness was completely mediated by maladaptive emotion regulation strategies, but not by stress sensitivity. However, in addition to maladaptive emotion regulation strategies, stress sensitivity demonstrated a mediation effect between ALE and attenuated positive psychotic positive symptoms when positive psychotic symptoms were grouped together. CONCLUSIONS: There are probably several possible trajectories leading to the formation of positive psychotic symptoms and the results of the present study reveal that one such trajectory may involve the maladaptive regulation of negative emotions alongside a certain general vulnerability after experiencing ALE.

The relationship between virtual body ownership and temperature sensitivity.

In the rubber hand illusion tactile stimulation seen on a rubber hand, that is synchronous with tactile stimulation felt on the hidden real hand, can lead to an illusion of ownership over the rubber hand. This illusion has been shown to produce a temperature decrease in the hidden hand, suggesting that such illusory ownership produces disownership of the real hand. Here we apply immersive virtual reality (VR) to experimentally investigate this with respect to sensitivity to temperature change. Forty participants experienced immersion in a VR with a virtual body (VB) seen from a first person perspective. For half the participants the VB was consistent in posture and movement with their own body, and in the other half there was inconsistency. Temperature sensitivity on the palm of the hand was measured before and during the virtual experience. The results show that temperature sensitivity decreased in the consistent compared to the inconsistent condition. Moreover, the change in sensitivity was significantly correlated with the subjective illusion of virtual arm ownership but modulated by the illusion of ownership over the full virtual body. This suggests that a full body ownership illusion results in a unification of the virtual and real bodies into one overall entity - with proprioception and tactile sensations on the real body integrated with the visual presence of the virtual body. The results are interpreted in the framework of a"body matrix" recently introduced into the literature.

Estimating extreme value cumulative distribution functions using bias-corrected kernel approaches

We propose a new kernel estimation of the cumulative distribution function based on transformation and on bias reducing techniques. We derive the optimal bandwidth that minimises the asymptotic integrated mean squared error. The simulation results show that our proposed kernel estimation improves alternative approaches when the variable has an extreme value distribution with heavy tail and the sample size is small.

A correlation sensitivity analysis of non-life underwriting risk in solvency capital requirement estimation

This paper analyses the impact of using different correlation assumptions between lines of business when estimating the risk-based capital reserve, the Solvency Capital Requirement -SCR-, under Solvency II regulations. A case study is presented and the SCR is calculated according to the Standard Model approach. Alternatively, the requirement is then calculated using an Internal Model based on a Monte Carlo simulation of the net underwriting result at a one-year horizon, with copulas being used to model the dependence between lines of business. To address the impact of these model assumptions on the SCR we conduct a sensitivity analysis. We examine changes in the correlation matrix between lines of business and address the choice of copulas. Drawing on aggregate historical data from the Spanish non-life insurance market between 2000 and 2009, we conclude that modifications of the correlation and dependence assumptions have a significant impact on SCR estimation.

Learning Preferences with Kernel-Based Methods

Learning of preference relations has recently received significant attention in machine learning community. It is closely related to the classification and regression analysis and can be reduced to these tasks. However, preference learning involves prediction of ordering of the data points rather than prediction of a single numerical value as in case of regression or a class label as in case of classification. Therefore, studying preference relations within a separate framework facilitates not only better theoretical understanding of the problem, but also motivates development of the efficient algorithms for the task. Preference learning has many applications in domains such as information retrieval, bioinformatics, natural language processing, etc. For example, algorithms that learn to rank are frequently used in search engines for ordering documents retrieved by the query. Preference learning methods have been also applied to collaborative filtering problems for predicting individual customer choices from the vast amount of user generated feedback. In this thesis we propose several algorithms for learning preference relations. These algorithms stem from well founded and robust class of regularized least-squares methods and have many attractive computational properties. In order to improve the performance of our methods, we introduce several non-linear kernel functions. Thus, contribution of this thesis is twofold: kernel functions for structured data that are used to take advantage of various non-vectorial data representations and the preference learning algorithms that are suitable for different tasks, namely efficient learning of preference relations, learning with large amount of training data, and semi-supervised preference learning. Proposed kernel-based algorithms and kernels are applied to the parse ranking task in natural language processing, document ranking in information retrieval, and remote homology detection in bioinformatics domain. Training of kernel-based ranking algorithms can be infeasible when the size of the training set is large. This problem is addressed by proposing a preference learning algorithm whose computation complexity scales linearly with the number of training data points. We also introduce sparse approximation of the algorithm that can be efficiently trained with large amount of data. For situations when small amount of labeled data but a large amount of unlabeled data is available, we propose a co-regularized preference learning algorithm. To conclude, the methods presented in this thesis address not only the problem of the efficient training of the algorithms but also fast regularization parameter selection, multiple output prediction, and cross-validation. Furthermore, proposed algorithms lead to notably better performance in many preference learning tasks considered.