• Publications
  • Influence
Use and misuse of population attributable fractions.
Computational and conceptual issues relevant to population attributable fraction estimation that are infrequently discussed elsewhere are considered, with illustrations from the breast cancer literature.
A log-linear approach to case-parent-triad data: assessing effects of disease genes that act either directly or through maternal effects and that may be subject to parental imprinting.
A log-linear method for analysis of case-parent-triad data, based on maximum likelihood with stratification on parental mating type, which generalizes easily to accommodate maternal effects on risk and produces powerful and orthogonal tests of the contribution of fetal versus maternal genetic factors.
Likelihood of conception with a single act of intercourse: providing benchmark rates for assessment of post-coital contraceptives.
It is found that the possibility of late ovulation produces a persistent risk of pregnancy even into the sixth week of the cycle, and post-coital contraceptives may be indicated even when intercourse has occurred late in the cycle.
Gene selection for sample classification based on gene expression data: study of sensitivity to choice of parameters of the GA/KNN method
The GA/KNN method is capable of selecting a subset of predictive genes from a large noisy data set for sample classification and is a multivariate approach that can capture the correlated structure in the data.
Residential Radon and Risk of Lung Cancer: A Combined Analysis of 7 North American Case-Control Studies
Direct evidence is provided of an association between residential radon and lung cancer risk, a finding predicted using miner data and consistent with results from animal and in vitro studies.
Non-hierarchical logistic models and case-only designs for assessing susceptibility in population-based case-control studies.
If the exposure and genetic categories occur independently and the disease is rare, then analyses based only on cases are valid, and offer better precision for estimating gene-environment interactions than those based on the full data.
Methods for detection of parent-of-origin effects in genetic studies of case-parents triads.
  • C. Weinberg
  • Biology, Medicine
    American journal of human genetics
  • 1 July 1999
This report reviews existing methods for detection of parent-of-origin effects, showing that each can be invalid under certain scenarios, and proposes new methods based on application of likelihood-based inference after stratification on both the parental mating type and the inherited number of copies of the allele under study.
Incidence of early loss of pregnancy.
The total rate of pregnancy loss after implantation, including clinically recognized spontaneous abortions, was 31 percent and most of the 40 women with unrecognized early pregnancy losses had normal fertility, since 95 percent of them subsequently became clinically pregnant within two years.
Designing and analysing case-control studies to exploit independence of genotype and exposure.
Maximum likelihood methods based on log-linear models that explicitly impose the independence assumption are described, something the usual logistic regression analyses cannot do.
Use of time to pregnancy to study environmental exposures.
To evaluate problems with collecting data on time to pregnancy, telephone interviews were conducted with nearly 700 pregnant women who reported having planned their pregnancies and power curves indicate that relatively small sample sizes are sufficient for investigating an exposure.