Nicolas Wicker

KD4v: comprehensible knowledge discovery system for missense variant

A major challenge in the post-genomic era is a better understanding of how human genetic alterations involved in disease affect the gene products. The KD4v (Comprehensible Knowledge Discovery System for Missense Variant) server allows to characterize and predict the phenotypic effects (deleterious/neutral) of missense variants. The server provides a set of rules learned by Induction Logic Programming (ILP) on a set of missense variants described by conservation, physico-chemical, functional and 3D structure predicates. These rules are interpretable by non-expert humans and are used to accurately predict the deleterious/neutral status of an unknown mutation. The web server is available at http://decrypthon.igbmc.fr/kd4v.

A new look towards BAC-based array CGH through a comprehensive comparison with oligo-based array CGH.

BackgroundCurrently, two main technologies are used for screening of DNA copy number; the BAC (Bacterial Artificial Chromosome) and the recently developed oligonucleotide-based CGH (Chromosomal Comparative Genomic Hybridization) arrays which are capable of detecting small genomic regions with amplification or deletion. The correlation as well as the discriminative power of these platforms has never been compared statistically on a significant set of human patient samples.ResultsIn this paper, we present an exhaustive comparison between the two CGH platforms, undertaken at two independent sites using the same batch of DNA from 19 advanced prostate cancers. The comparison was performed directly on the raw data and a significant correlation was found between the two platforms. The correlation was greatly improved when the data were averaged over large chromosomic regions using a segmentation algorithm. In addition, this analysis has enabled the development of a statistical model to discriminate BAC outliers that might indicate microevents. These microevents were validated by the oligo platform results.ConclusionThis article presents a genome-wide statistical validation of the oligo array platform on a large set of patient samples and demonstrates statistically its superiority over the BAC platform for the Identification of chromosomic events. Taking advantage of a large set of human samples treated by the two technologies, a statistical model has been developed to show that the BAC platform could also detect microevents.

A maximum likelihood approximation method for Dirichlet's parameter estimation

Dirichlet distributions are natural choices to analyse data described by frequencies or proportions since they are the simplest known distributions for such data apart from the uniform distribution. They are often used whenever proportions are involved, for example, in text-mining, image analysis, biology or as a prior of a multinomial distribution in Bayesian statistics. As the Dirichlet distribution belongs to the exponential family, its parameters can be easily inferred by maximum likelihood. Parameter estimation is usually performed with the Newton-Raphson algorithm after an initialisation step using either the moments or Ronnings methods. However this initialisation can result in parameters that lie outside the admissible region. A simple and very efficient alternative based on a maximum likelihood approximation is presented. The advantages of the presented method compared to two other methods are demonstrated on synthetic data sets as well as for a practical biological problem: the clustering of protein sequences based on their amino acid compositions.

Minimal enclosing parallelepiped in 3D

We investigate the problem of finding a minimal volume parallelepiped enclosing a given set of n three-dimensional points. We give two mathematical properties of these parallelepipeds, from which we derive two algorithms of theoretical complexity O(n6). Experiments show that in practice our quickest algorithm runs in O(n2) (at least for n ≤ 105). We also present our application in structural biology.

℮-conome: an automated tissue counting platform of cone photoreceptors for rodent models of retinitis pigmentosa

BackgroundRetinitis pigmentosa is characterized by the sequential loss of rod and cone photoreceptors. The preservation of cones would prevent blindness due to their essential role in human vision. Rod-derived Cone Viability Factor is a thioredoxin-like protein that is secreted by rods and is involved in cone survival. To validate the activity of Rod-derived Cone Viability Factors (RdCVFs) as therapeutic agents for treating retinitis Pigmentosa, we have developed e-conome, an automated cell counting platform for retinal flat mounts of rodent models of cone degeneration. This automated quantification method allows for faster data analysis thereby accelerating translational research.MethodsAn inverted fluorescent microscope, motorized and coupled to a CCD camera records images of cones labeled with fluorescent peanut agglutinin lectin on flat-mounted retinas. In an average of 300 fields per retina, nine Z-planes at magnification X40 are acquired after two-stage autofocus individually for each field. The projection of the stack of 9 images is subject to a threshold, filtered to exclude aberrant images based on preset variables. The cones are identified by treating the resulting image using 13 variables empirically determined. The cone density is calculated over the 300 fields.ResultsThe method was validated by comparison to the conventional stereological counting. The decrease in cone density in rd1 mouse was found to be equivalent to the decrease determined by stereological counting. We also studied the spatiotemporal pattern of the degeneration of cones in the rd1 mouse and show that while the reduction in cone density starts in the central part of the retina, cone degeneration progresses at the same speed over the whole retinal surface. We finally show that for mice with an inactivation of the Nucleoredoxin-like genes Nxnl1 or Nxnl2 encoding RdCVFs, the loss of cones is more pronounced in the ventral retina.ConclusionThe automated platform ℮-conome used here for retinal disease is a tool that can broadly accelerate translational research for neurodegenerative diseases.

Multidimensional Fitting for Multivariate Data Analysis

Large multidimensional data matrices are frequent in biology. However, statistical methods often have difficulties dealing with such matrices because they contain very complex data sets. Consequently variable selection and dimensionality reduction methods are often used to reduce matrix complexity, although at the expense of information conservation. A new method derived from multidimensional scaling (MDS) is presented for the case where two matrices are available to describe the same population. The presented method transforms one of the matrices, called the target matrix, with some constraints to make it fit with the second matrix, referred to as the reference matrix. The fitting to the reference matrix is performed on the distances computed for the two matrices, and the transformation depends on the problem at hand. A special feature of this method is that a variable can be only partially modified. The method is applied on the exclusive-or (XOR) problem and then on a biological application with large-scale gene expression data.