Incorporating Concomitant Medications into Genome-Wide Analyses for the Study of Complex Disease and Drug Response.

Given the high costs of conducting a drug-response trial, researchers are now aiming to use retrospective analyses to conduct genome-wide association studies (GWAS) to identify underlying genetic contributions to drug-response variation. To prevent confounding results from a GWAS to investigate drug response, it is necessary to account for concomitant medications, defined as any medication taken concurrently with the primary medication being investigated. We use data from the Action to Control Cardiovascular Disease (ACCORD) trial in order to implement a novel scoring procedure for incorporating concomitant medication information into a linear regression model in preparation for GWAS. In order to accomplish this, two primary medications were selected: thiazolidinediones and metformin because of the wide-spread use of these medications and large sample sizes available within the ACCORD trial. A third medication, fenofibrate, along with a known confounding medication, statin, were chosen as a proof-of-principle for the scoring procedure. Previous studies have identified SNP rs7412 as being associated with statin response. Here we hypothesize that including the score for statin as a covariate in the GWAS model will correct for confounding of statin and yield a change in association at rs7412. The response of the confounded signal was successfully diminished from p = 3.19 × 10-7 to p = 1.76 × 10-5, by accounting for statin using the scoring procedure presented here. This approach provides the ability for researchers to account for concomitant medications in complex trial designs where monotherapy treatment regimens are not available.

Use of Bayesian Inference in Crystallographic Structure Refinement via Full Diffraction Profile Analysis.

A Bayesian inference method for refining crystallographic structures is presented. The distribution of model parameters is stochastically sampled using Markov chain Monte Carlo. Posterior probability distributions are constructed for all model parameters to properly quantify uncertainty by appropriately modeling the heteroskedasticity and correlation of the error structure. The proposed method is demonstrated by analyzing a National Institute of Standards and Technology silicon standard reference material. The results obtained by Bayesian inference are compared with those determined by Rietveld refinement. Posterior probability distributions of model parameters provide both estimates and uncertainties. The new method better estimates the true uncertainties in the model as compared to the Rietveld method.

Scientists urge DHS to improve Bioterrorism Risk Assessment.

In 2006, the Department of Homeland Security (DHS) completed its first Bioterrorism Risk Assessment (BTRA), intended to be the foundation for DHS's subsequent biennial risk assessments mandated by Homeland Security Presidential Directive 10 (HSPD-10). At the request of DHS, the National Research Council established the Committee on Methodological Improvements to the Department of Homeland Security's Biological Agent Risk Analysis to provide an independent, scientific peer review of the BTRA. The Committee found a number of shortcomings in the BTRA, including a failure to consider terrorists as intelligent adversaries in their models, unnecessary complexity in threat and consequence modeling and simulations, and a lack of focus on risk management. The Committee unanimously concluded that an improved BTRA is needed to provide a more credible foundation for risk-informed decision making.