Robust Selection of the Most Discriminative Variables in the Dichotomous Problem with Application to Some Respiratory Disease Data

A robust method for selection of variables with the greatest discriminatory power is presented in the paper. The method deals with the two groups of data problem. An application of the method to some respiratory disease data and comparisons with classical procedures are given, also.     

Reference values for feeding parameters of isopods (Porcellio scaber,Isopoda, Crustacea)

The advantage of using terrestrial isopods in toxicity studies is that a battery of parameters can be tested at different levels of biological complexity. Feeding parameters for example link organism level response to potential ecological consequences but a problem with using feeding parameters in toxicity tests with terrestrial isopods is their high variability. The aim of our study was to set benchmark values for feeding parameters for isopod Porcellio scaber (Isopoda, Crustacea) in laboratory-controlled experiments. In the work presented here, the daily feeding rate of the central 50% of the control population of Porcellio scaber and a correlation between feeding rate and isopod weight were set. Values outside these ranges need additional evaluation to increase the relevance of test outcomes. We suggest using benchmark values for feeding parameters as well as the coefficient of variation (a) to identify animals with altered feeding parameters with respect to controls, and (b) to assess the data quality in each experiment.     

Anatomically-based skeletal coordinate systems for use with impact biomechanics data intended for anthropomorphic test device development

Post-mortem human subjects (PMHS) are frequently used to characterize biomechanical response and injury tolerance of humans to various types of loading by means of instrumentation installed directly on the skeleton. Data extracted from such tests are often used to develop and validate anthropomorphic test devices (ATDs), which function as human surrogates in tests for injury assessment. Given that the location and orientation of installed instrumentation differs between subjects, nominally similar measurements made on different PMHS must be transformed to standardized, skeletal-based local coordinate systems (LCS) before appropriate data comparisons can be made. Standardized PMHS LCS that correspond to ATD instrumentation locations and orientations have not previously been published. This paper introduces anatomically-defined PMHS LCS for body regions in which kinematic measurements are made using ATDs. These LCS include the head, sternum, single vertebrae, pelvis, femurs (distal and proximal), and tibiae (distal and proximal) based upon skeletal landmarks extracted from whole body CT scans. The proposed LCS provide a means to standardize the reporting of PMHS data, and facilitate both the comparison of PMHS impact data across institutions and the application of PMHS data to the development and validation of ATDs.     

An Algorithm for Calculation of n-Dimensional Nonlinear Dependencies – Behaviour Tests on Rats

The evaluation of the data obtained during the behaviour tests always leads to the problem of multiple correlation, very often with non-linear dependencies on the target. All mathematical and statistical procedures that have been used so far are based on the assumption of an equation for the desired correlation for which parameters and related statistical equivalents are determined eventually. The MODAK system applied here (MODAK = algorithms of modelling for the calculation of multi-dimensional non-linear mathematical models) breaks down a complex correlation into individual dependencies in a mathematical and statistical way and selects suitable equations for each of them independently and determines the corresponding parameters. The numerical example evaluates data of behaviour tests on rats. First results obtained on the correlations of various behaviour tests indicate both the possibility of selecting suitable tests independent of each other and a better interpretation of the observed patterns of behaviour taking into account the interrelations between the tests. In addition, MODAK is a method which can be applied as a matter of course in a general way to all cases which call for the reduction and analysis of data occurring in process and system analysis and in the evaluation of test results requiring statistical modelling. So far, MODAK applications range from engineering sciences to medicine.     

Assessment of Optimal Selected Prognostic Factors

The identification and assessment of prognostic factors is one of the major tasks in clinical research. The assessment of one single prognostic factor can be done by recently established methods for using optimal cutpoints. Here, we suggest a method to consider an optimal selected prognostic factor from a set of prognostic factors of interest. This can be viewed as a variable selection method and is the underlying decision problem at each node of various tree building algorithms. We propose to use maximally selected statistics where the selection is defined over the set of prognostic factors and over all cutpoints in each prognostic factor. We demonstrate that it is feasible to compute the approximate null distribution. We illustrate the new variable selection test with data of the German Breast Cancer Study Group and of a small study on patients with diffuse large B‐cell lymphoma. Using the null distribution for a p‐value adjusted regression trees algorithm, we adjust for the number of variables analysed at each node as well. (© 2004 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Progeny testing for recessive genes: procedures and interpretations

Summary Methods for calculating the probability of detecting a carrier of a recessive gene by utilizing matings among related individuals are presented for single and litter bearing species. The confidence level for detection of heterozygosity depends upon: (1) the genetic relationship between mates, (2) the number of mates per male and the number of offspring per mate, (3) whether an estimate of recessive gene frequency before selection is available and (4) the magnitude of that frequency. Methods of computing probability of heterozygosity vs homozygosity utilizing Bayes theorem also are presented. In the conventional progeny test method, a sire initially is assumed heterozygous before calculations are made, but no prior information concerning his probable genotype is utilized. In the method using Bayes theorem, prior sources of information from relatives or from estimates of population allele frequency are utilized. This method gives the exact probability that a sire is not a carrier, given prior information and that he produces all normal offspring. These methods could be used in any sexually reproducing species to identify not only detrimental genes but beneficial genes as well.     

Statistical methods for model comparison in parameter estimation problems for distributed systems

In this note we outline some recent results on the development of a statistical testing methodology for inverse problems involving partial differential equation models. Applications to several problems from biology are presented. The statistical tests, which are in the spirit of analysis of variance (ANOVA), are based on asymptotic distributional results for estimators and residuals in a least squares approach.Research supported in part under grants NSF MCS 8504316, NASA NAG-1-517, and AFOSRF-49620-86-C-0111. Part of this research was carried out while the first author was a visiting scientist at the Institute for Computer Applications in Science and Engineering (ICASE), NASA Langley Research Center, Hampton, VA, which is operated under NASA contracts NASI-18107 and NASI-18605     

Testing some hypotheses on human evolution and sexual dimorphism

Research on human evolution and sexual dimorphism motivates an interesting test problem. In studying hominid phylogeny it is of interest to test whether parallel evolution plays a role. With regard to sexual dimorphism it is of interest to known whether the directions of sexual dimorphism in the populations being compared are the same. We show that testing these two problems gives rise to the same type of hypothesis testing, viz. the problem of testing the hypothesis that the means of independent, normally distributed random vectors with unit covariance matrices are situated on a straight line through the origin. A test is proposed and applied to study the sexual dimorphism of 20 recent skull populations. In this example the hypothesis of equal directions of sexual dimorphism is rejected. The classical theory of constructing multiple discriminant functions (canonical variates) is adapted to the problem of comparing sexual dimorphisms.     

A practical test for in vitro evaluation of photosensitization: Assessment with hematoporphyrin derivative (HPD)

A simple procedure is described for the determination of the photosensitizing potency of drugs, using three leukemic cell lines, two of lymphocytic origin, L1210 and P388 and one of erythroid type, Friend-745. The procedure allows one to investigate several aspects of the photosensitization properties of tested compounds such as cellular localization and direct (trypan blue exclusion) or delayed (clonogenicity) photomediated toxicities.The method was assessed using crude hematoporphyrin derivative (HPD) as well as dihematoporphyrin ether (DHE) or commercially available Photofrin II. Results were compared to those obtained with normal cells, e.g spleen lymphocytes and erythropoietic stem cells (CFU-e), and discussed in the light of the relative response of normal versus transformed cells.Abbreviations DHE Dihematoporphyrin Ether - FCS Fetal Calf Serum - HPD Hematoporphyrin Derivative - PDT Photodynamic Therapy