Cryptic multiple hypotheses testing in linear models: overestimated effect sizes and the winner's curse

Fitting generalised linear models (GLMs) with more than one predictor has become the standard method of analysis in evolutionary and behavioural research. Often, GLMs are used for exploratory data analysis, where one starts with a complex full model including interaction terms and then simplifies by removing non-significant terms. While this approach can be useful, it is problematic if significant effects are interpreted as if they arose from a single a priori hypothesis test. This is because model selection involves cryptic multiple hypothesis testing, a fact that has only rarely been acknowledged or quantified. We show that the probability of finding at least one ‘significant’ effect is high, even if all null hypotheses are true (e.g. 40% when starting with four predictors and their two-way interactions). This probability is close to theoretical expectations when the sample size (N) is large relative to the number of predictors including interactions (k). In contrast, type I error rates strongly exceed even those expectations when model simplification is applied to models that are over-fitted before simplification (low N/k ratio). The increase in false-positive results arises primarily from an overestimation of effect sizes among significant predictors, leading to upward-biased effect sizes that often cannot be reproduced in follow-up studies (‘the winner's curse’). Despite having their own problems, full model tests and P value adjustments can be used as a guide to how frequently type I errors arise by sampling variation alone. We favour the presentation of full models, since they best reflect the range of predictors investigated and ensure a balanced representation also of non-significant results.     

Fetal de novo heterozygous variant in the isocitrate dehydrogenase 1 gene associated with growth restriction,skeletal, cerebral and vascular anomalies

Germline pathogenic variants in isocitrate dehydrogenase 1 (IDH1) can lead to a rare neurodevelopmental disorder called metaphyseal chondromatosis with D-2-hydroxyglutaric aciduria, including severe skeletal and cerebral anomalies. To the best of our knowledge, no prenatal case of an IDH1 pathogenic variant has been reported in literature. Somatic sequence variants in IDH1/2 genes are described in distinct cancers, premalignant diseases and rare inherited metabolic disorders. Amniocentesis and further genetic testing including trio exome sequencing were performed due to suspicious findings on a second trimester routine prenatal ultrasound examination. The fetus was found to have growth restriction, cerebral abnormalities (ex vacuo hydrocephalus, cerebellar and vermian hypoplasia, corpus callosum dysgenesis), brachycephaly, narrow chest, persistent left superior vena cava, liver calcifications, hyperechogenic bowel, short tubular bones and joint contractures. A de novo heterozygous variant in the IDH1 gene was detected via trio exome sequencing. The prenatal diagnosis of a de novo pathogenic variant in IDH1 in a fetus with the described phenotype, obtained through trio exome sequencing, helped parents and providers with an informed decision making about pregnancy management.     

Expositionsabschätzung mittels Modellrechnungen

Human exposure to polycyclic aromatic hydrocarbons (PAH) via daily liquid consumption (primarily through the intake of drinking water) is estimated using model calculations based on the indicator substance benzo(a)pyrene. The additional contribution to oral exposure by a consumption of solid nutrients, and by soil and dust ingestion, are evaluated using typical background conditions in Germany (conurbations of North-Rhine-Westphalia). A series of further scenarios is then applied to examine the influence of elevated PAH concentrations in the daily intake of liquid. The degree of PAH exposure via liquid only equals or exceeds the exposure via the solid fraction of the diet under very specific conditions. The contribution via liquid consumption to a total lifelong exposure of both BAP and PAH caused by dietary intake may be estimated as approximately 4 %. It increases to 56 % under a level equivalent to the Dutch intervention value for groundwater and reaches 90 % under a level equivalent to the WHO guideline for drinking water. Considering occupants who live on PAH contaminated garden allotments with private wells, the highest degree of exposure is caused by vegetables followed by house dust, then liquid diet and finally by soil. Finally, the model approach is used for evaluating the success of potential remedial treatments with respect to exposure reduction. In addition to the application of average values, conservative estimates of exposure in terms of 95 % values are set out in this publication based on sumulation methodologies.     

Non-invasive prenatal detection of achondroplasia in size-fractionated cell-free DNA by MALDI-TOF MS assay

Achondroplasia is the most common form of short-limbed dwarfism in humans and is caused by mutations in the FGFR3 gene. Currently, prenatal diagnosis of this disorder relies on invasive procedures. Recent studies have shown that fetal single gene point mutations could be detected in cell-free DNA (cf-DNA) from maternal plasma by either the matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) assay with single allele base extension reaction (SABER) approach or the size fractionation of cf-DNA in maternal plasma. Here, we combined the two approaches to non-invasively examine the fetal G1138A mutation in maternal plasma. cf-DNA was extracted from maternal plasma samples obtained from two pregnant women at risk for achondroplasia. The fetal G1138A mutation was determined by the analysis of size-fractionated cf-DNA in maternal plasma using MALDI-TOF MS with SABER approach and homogenous MassEXTEND (hME) assay, respectively. The fetal G1138A mutation was detectable in the two achondroplasia-affected pregnancies by the analysis of cf-DNA in maternal plasma using MALDI-TOF MS. However, the size-fractionation approach led to a more precise detection of the fetal mutation in both analyses. This analysis would be suitable for non-invasive prenatal diagnosis of diseases caused by fetal single gene point mutations. Copyright © 2007 John Wiley & Sons, Ltd.     

Simulating discrete event systems with UML and JAVA

Summary  The main readership are students in the field of informatics. In most parts, other interested scientists from environmental disciplines get some insights in the field of environmental modelling and software. The book can be recommended broadly. With the last chapter ‘Simulation in Practice’, this work may also be useful for commercial application. With chapter 14, the authors contribute to the increasingly important topic of E-Learning.