van Iterson M.,Leiden University |
van Iterson M.,Netherlands Bioinformatics Center |
Boer J.M.,Leiden University |
Boer J.M.,Netherlands Bioinformatics Center |
And 4 more authors.
BMC Bioinformatics | Year: 2010
Background: In high-dimensional data analysis such as differential gene expression analysis, people often use filtering methods like fold-change or variance filters in an attempt to reduce the multiple testing penalty and improve power. However, filtering may introduce a bias on the multiple testing correction. The precise amount of bias depends on many quantities, such as fraction of probes filtered out, filter statistic and test statistic used.Results: We show that a biased multiple testing correction results if non-differentially expressed probes are not filtered out with equal probability from the entire range of p-values. We illustrate our results using both a simulation study and an experimental dataset, where the FDR is shown to be biased mostly by filters that are associated with the hypothesis being tested, such as the fold change. Filters that induce little bias on the FDR yield less additional power of detecting differentially expressed genes. Finally, we propose a statistical test that can be used in practice to determine whether any chosen filter introduces bias on the FDR estimate used, given a general experimental setup.Conclusions: Filtering out of probes must be used with care as it may bias the multiple testing correction. Researchers can use our test for FDR bias to guide their choice of filter and amount of filtering in practice. © 2010 van Iterson et al; licensee BioMed Central Ltd. Source
Amin N.,Erasmus Medical Center |
Schuur M.,Erasmus Medical Center |
Gusareva E.S.,Academy of Sciences of the Czech Republic |
Isaacs A.,Erasmus Medical Center |
And 9 more authors.
Molecular Psychiatry | Year: 2012
The NEO-Five-Factor Inventory divides human personality traits into five dimensions: neuroticism, extraversion, openness, conscientiousness and agreeableness. In this study, we sought to identify regions harboring genes with large effects on the five NEO personality traits by performing genome-wide linkage analysis of individuals scoring in the extremes of these traits (90th percentile). Affected-only linkage analysis was performed using an Illumina 6K linkage array in a family-based study, the Erasmus Rucphen Family study. We subsequently determined whether distinct, segregating haplotypes found with linkage analysis were associated with the trait of interest in the population. Finally, a dense single-nucleotide polymorphism genotyping array (Illumina 318K) was used to search for copy number variations (CNVs) in the associated regions. In the families with extreme phenotype scores, we found significant evidence of linkage for conscientiousness to 20p13 (rs1434789, log of odds (LOD)5.86) and suggestive evidence of linkage (LOD 2.8) for neuroticism to 19q, 21q and 22q, extraversion to 1p, 1q, 9p and12q, openness to 12q and 19q, and agreeableness to 2p, 6q, 17q and 21q. Further analysis determined haplotypes in 21q22 for neuroticism (P-values0.009, 0.007), in 17q24 for agreeableness (marginal P-value0.018) and in 20p13 for conscientiousness (marginal P-values0.058, 0.038) segregating in families with large contributions to the LOD scores. No evidence for CNVs in any of the associated regions was found. Our findings imply that there may be genes with relatively large effects involved in personality traits, which may be identified with next-generation sequencing techniques. © 2012 Macmillan Publishers Limited All rights reserved. Source
Schol-Gelok S.,Erasmus University Rotterdam |
Janssens A.C.J.W.,Erasmus University Rotterdam |
Tiemeier H.,Erasmus University Rotterdam |
Liu F.,Erasmus University Rotterdam |
And 10 more authors.
Biological Psychiatry | Year: 2010
Background: Depression has a strong genetic component but candidate gene studies conducted to date have not shown consistent associations. Methods: We conducted a genome-wide parametric and nonparametric linkage analysis in a large-scale family-based study including 115 individuals with depression who were identified based on the Hospital Anxiety Depression Scale, Center for Epidemiologic Studies Depression Rating Scale, or use of antidepressive medication. Further, we investigated the most promising chromosomal regions found in the genome-wide linkage analysis with an association analysis in 734 individuals in the family-based study and 2373 individuals in the population-based study. Results: Our study demonstrated evidence for significant linkage of depression to chromosome 2p16.1-15 (logarithm of odds [LOD]= 5.13; parametric analysis) and suggestive evidence for linkage in nonparametric analysis to chromosome 5p15.33 (LOD=2.14), 11q25 (LOD=2.27), and 19p13.3 (LOD=2.66). The subsequent association analysis in the family-based study showed region-wide significant association in intron 1 of the OPCML gene on chromosome 11q25 (empirical p value=.04). The association analysis in the population-based study did not show any region-wide significant association, yet showed suggestive association in intron 1 of the APLP2 gene on chromosome 11q25. Conclusions: Our linkage and association studies suggest a locus for depression on chromosomes 2p16.1-15 and 11q25. The linkage to chromosome 11q25 may be, in part, explained by the OPCML or the APLP2 gene. Further, there is evidence for a role of the GNG7 gene (chromosome 19p13.3). © 2010 Society of Biological Psychiatry. Source
Rigter T.,VU University Amsterdam |
van Aart C.J.A.,VU University Amsterdam |
Elting M.W.,VU University Amsterdam |
Waisfisz Q.,VU University Amsterdam |
And 4 more authors.
Clinical Genetics | Year: 2014
Next-generation sequencing is increasingly being chosen as a diagnostic tool for cases of expected genetic, but unresolved origin. The consequential increased need for decisions on disclosure of unsolicited findings poses a challenge for the informed consent procedure. This study explored the first experiences with, and needs for, the informed consent procedure in diagnostic exome sequencing, with the stakeholders involved. Semi-structured interviews were conducted with 11 professional experts and one professional gave a written response. Furthermore, the counseling process was observed in three cases where exome sequencing was offered, followed by interviews with the patient (representative) and the genetic counselor. The respondents not only preferred an opt-out for unsolicited findings but also identified many challenges and therefore more experiences with exome sequencing was considered needed. Context-dependent decision-making was observed and an Advisory Board for unsolicited findings was considered helpful while doubts were raised about the feasibility and the possibility of undermining patients' autonomy. Finally, respondents brought up the complexity of information provision, and division of responsibilities between clinicians and the lab. These challenges and needs, raised by stakeholders involved, provide more insight in the next steps needed for an optimal informed consent procedure for exome sequencing in diagnostics. © 2013 The Authors. Clinical Genetics published by JohnWiley & Sons A/S. Published by John Wiley & Sons Ltd. Source
Brouwers M.C.G.J.,Maastricht University |
Konrad R.J.,Eli Lilly and Company |
van Himbergen T.M.,Tufts University |
Isaacs A.,Erasmus Medical Center |
And 8 more authors.
Nutrition, Metabolism and Cardiovascular Diseases | Year: 2013
Background and aims: Two recent independent studies showed that patients with familial combined hyperlipidemia (FCHL) have elevated plasma levels of proprotein convertase subtilisin kexin type 9 (PCSK9) and markers of cholesterol synthesis. Both PCSK9 expression and cholesterol synthesis are downstream effects of hepatic activation of sterol regulatory element binding protein 2 (SREBP2). The present study was conducted to study the relationship between plasma PCSK9 and markers of cholesterol synthesis in FCHL. Methods and results: Markers of cholesterol synthesis (squalene, desmosterol, lathosterol), cholesterol absorption (campesterol, sitosterol, cholestanol) and PCSK9 were measured in plasma of FCHL patients (n=103) and their normolipidemic relatives (NLR; n=240). Plasma PCSK9, lathosterol and desmosterol levels were higher in FCHL patients than their NLR (p<0.001, age and sex adjusted). Heritability calculations demonstrated that 35% of the variance in PCSK9 levels could be explained by additive genetic effects (p<0.001). Significant age- and sex-adjusted correlations were observed for the relationship between PCSK9 and lathosterol, both unadjusted and adjusted for cholesterol, in the overall FCHL population (both p<0.001). Multivariate regression analyses, with PCSK9 as the dependent variable, showed that the regression coefficient for FCHL status decreased by 25% (from 0.8 to 0.6) when lathosterol was included. Nevertheless, FCHL status remained an independent contributor to plasma PCSK9 (p<0.001). Conclusions: The present study confirms the previously reported high and heritable PCSK9 levels in FCHL patients. Furthermore, we now show that high PCSK9 levels are, in part, explained by plasma lathosterol, suggesting that SREBP2 activation partly accounts for elevated PCSK9 levels in FCHL. © 2012 Elsevier B.V. Source