Component-wise gradient boosting and false discovery control in survival analysis with high-dimensional covariates.

MOTIVATION: Technological advances that allow routine identification of high-dimensional risk factors have led to high demand for statistical techniques that enable full utilization of these rich sources of information for genetics studies. Variable selection for censored outcome data as well as control of false discoveries (i.e. inclusion of irrelevant variables) in the presence of high-dimensional predictors present serious challenges. This article develops a computationally feasible method based on boosting and stability selection. Specifically, we modified the component-wise gradient boosting to improve the computational feasibility and introduced random permutation in stability selection for controlling false discoveries. RESULTS: We have proposed a high-dimensional variable selection method by incorporating stability selection to control false discovery. Comparisons between the proposed method and the commonly used univariate and Lasso approaches for variable selection reveal that the proposed method yields fewer false discoveries. The proposed method is applied to study the associations of 2339 common single-nucleotide polymorphisms (SNPs) with overall survival among cutaneous melanoma (CM) patients. The results have confirmed that BRCA2 pathway SNPs are likely to be associated with overall survival, as reported by previous literature. Moreover, we have identified several new Fanconi anemia (FA) pathway SNPs that are likely to modulate survival of CM patients. AVAILABILITY AND IMPLEMENTATION: The related source code and documents are freely available at https://sites.google.com/site/bestumich/issues. CONTACT: yili@umich.edu.

Variations of chlorophyll-a in the northeastern Indian Ocean after the 2004 South Asian tsunami

Analysis of satellite remote sensing data has revealed changes in distribution of chlorophyll-a (Chl-a) and sea surface temperature (SST) in the Indian Ocean during the South Asian tsunami in December 2004. Chl-a data derived from Moderate Resolution Imaging Spectroradiometer (MODIS) and Sea-viewing Wide Field-ofview Sensor (SeaWiFS) images were examined for the period from 1998 to 2005. Around the epicentre of the Sumatra earthquake, the Chl-a concentrationwas found to increase prior to the main event on 26 December 2004 and then decrease during the tsunami event, while a high SST (~30-31°C) was observed in and around the epicentral region. Chl-a concentrations in the coastal waters of the Southeast Asian countries were remarkably low during and after the tsunami. Similar but relatively small variations inChl-a and SST were observed during the second earthquake on 28 March 2005. Analysis of Chl-a, SST, wind and upwelling water has provided information for understanding the changes in Chl-a concentration during the tsunami. A very large offshore phytoplankton bloom (~300 km2) appeared to the southeast of Sri Lanka about 3 weeks after the tsunami; this might have been caused by a tropical storm that could be responsible for the enhancement of nutrients. © 2009 Taylor & Francis.

Droplet oscillations in high gradient magnetic field

In high intensity and high gradient magnetic fields the volumetric force on diamagnetic material, such as water, leads to conditions very similar to microgravity in a terrestrial laboratory. In principle, this opens the possibility to determine material properties of liquid samples without wall contact, even for electrically non-conducting materials. In contrast, AC field levitation is used for conductors, but then terrestrial conditions lead to turbulent flow driven by Lorentz forces. DC field damping of the flow is feasible and indeed practiced to allow property measurements. However, the AC/DC field combination acts preferentially on certain oscillation modes and leads to a shift in the droplet oscillation spectrum.What is the cause? A nonlinear spectral numerical model is presented, to address these problems

Droplet oscillations in high gradient static magnetic field

In high intensity and high gradient magnetic fields the volumetric force on diamagnetic material, such as water, leads to conditions very similar to microgravity in a terrestrial laboratory. In principle, this opens the possibility to determine material properties of liquid samples without wall contact, even for electrically non-conducting materials. In contrast, AC field levitation is used for conductors, but then terrestrial conditions lead to turbulent flow driven by Lorentz forces. DC field damping of the flow is feasible and indeed practiced to allow property measurements. However, the AC/DC field combination acts preferentially on certain oscillation modes and leads to a shift in the droplet oscillation spectrum.What is the cause? A nonlinear spectral numerical model is presented, to address these problems.