COMPARISM OF VARIABLE SELECTION TECHNIQUES IN REGRESSION OF MICROARRAY DATA

$33,899P20FY2009RRNIH

University Of Vermont & St Agric College, Burlington VT

Investigators

Linked publications & trials

Paper 33724001 Paper 30369837 Paper 30220884 Paper 29414025 Paper 28944500 Paper 27381916 Paper 26794592 Paper 26703714 Paper 26606276 Paper 26175502 Paper 25969542 Paper 25897958 Paper 25886764 Paper 25562295 Paper 25520806 Paper 25401169 Paper 25351596 Paper 25309735 Paper 25061893 Paper 24813700 Paper 24576858 Paper 24317525 Paper 24268632 Paper 23800453 Paper 23608191 Paper 23597827 Paper 23364479 Paper 23334188 Paper 23311620 Paper 23275028 Paper 23178720 Paper 23157360 Paper 23146917 Paper 23124522 Paper 23098169 Paper 22982281 Paper 22967899 Paper 22910360 Paper 22827337 Paper 22810990 Paper 22807455 Paper 22720940 Paper 22689812 Paper 22595869 Paper 22459246 Paper 22450950 Paper 22434832 Paper 22394466 Paper 22348807 Paper 22300531 Paper 22249911 Paper 22246506 Paper 22246505 Paper 22228505 Paper 22091532 Paper 22083608 Paper 22028429 Paper 21998646 Paper 21978630 Paper 21961592 Paper 21879738 Paper 21875652 Paper 21866565 Paper 21850205 Paper 21782899 Paper 21778411 Paper 21704744 Paper 21667757 Paper 21635816 Paper 21627827 Paper 21602374 Paper 21590759 Paper 21538880 Paper 21517145 Paper 21505409 Paper 21448062 Paper 21437602 Paper 21419341 Paper 21399873 Paper 21376016 Paper 21318207 Paper 21295035 Paper 21222438 Paper 21209930 Paper 21178106 Paper 21172408 Paper 21049007 Paper 20980644 Paper 20933223 Paper 20845963 Paper 20817764 Paper 20728429 Paper 20718491 Paper 20715768 Paper 20681608 Paper 20616001 Paper 20603077 Paper 20566348 Paper 20537404 Paper 20506281

Abstract

This subproject is one of many research subprojects utilizing the resources provided by a Center grant funded by NIH/NCRR. The subproject and investigator (PI) may have received primary funding from another NIH source, and thus could be represented in other CRISP entries. The institution listed is for the Center, which is not necessarily the institution for the investigator. The selection of relevant predictor variables in regression is crucial to have an accurate predictive model. DNA microarray data consists of thousands of genes as predictor variables that greatly exceed the number of observations in an experiment. It is advantageous to group highly correlated variables to simplify and improve the predictive model. The overall goal of this research is to compare three techniques used to find the optimal grouping of related genes for regression: hierarchical clustering, lasso, and elastic net. In addition, two outcome predictive models will be compared. Modeling and predicting changes at the DNA level will assist in understanding the development of complex diseases. Reducing the number of variables in the model will allow for an easier interpretation of results.

View original record on NIH RePORTER →