Variable selection for qualitative interactions in personalized medicine while controlling the family-wise error rate.

For many years, subset analysis has been a popular topic for the biostatistics and clinical trials literature. In more recent years, the discussion has focused on finding subsets of genomes which play a role in the effect of treatment, often referred to as stratified or personalized medicine. Though highly sought after, methods for detecting subsets with altering treatment effects are limited and lacking in power. In this article we discuss variable selection for qualitative interactions with the aim to discover these critical patient subsets. We propose a new technique designed specifically to find these interaction variables among a large set of variables while still controlling for the number of false discoveries. We compare this new method against standard qualitative interaction tests using simulations and give an example of its use on data from a randomized controlled trial for the treatment of depression.

Methods to determine the minimum important difference for a sexual event diary used by postmenopausal women with hypoactive sexual desire disorder.

INTRODUCTION: Recently, there has been much discussion in the literature about how to determine the meaningfulness of results generated from a patient-reported outcome measure. A number of reviews have shown that there are two main approaches: anchor- and distribution-based approaches for determining the minimum important difference (MID) for a new measure. There are issues with calculating an MID using each method: Will the two approaches give the same estimate? If the estimates differ, how do you decide on one estimate? Would asking patients directly be more beneficial? AIM: A case study was presented to address these issues based on a newly developed diary assessing number of satisfactory sexual events (SSEs) per week in women with hypoactive sexual desire disorder (HSDD). METHODS: Anchor- and distribution-based estimates were generated from data gathered in two double-blind, placebo-controlled, parallel group trials for the treatment of HSDD (N = 788). A novel interview study was used to ask women directly about an MID for SSEs (N = 77). MAIN OUTCOME MEASURES: Defining the MID for an SSE diary in women with HSDD. RESULTS: The estimates varied, producing a range of mean MID estimates between 0.04 and 0.46 SSEs per week. CONCLUSION: We recommend that rather than defining the MID, a range should be selected from the set of estimates formed by the limits of the 95% confidence intervals.

Efficient computation and model selection for the support vector regression.

In this letter, we derive an algorithm that computes the entire solution path of the support vector regression (SVR). We also propose an unbiased estimate for the degrees of freedom of the SVR model, which allows convenient selection of the regularization parameter.