For Manuscript Submission, Check or Review Login please go to Submission Websites List.
For the academic login, please select your country in the dropdown list. You will be redirected to verify your credentials.
Detecting Rare Variants for Quantitative Traits Using Nuclear FamiliesGuo W.a, b · Shugart Y.Y.a
aDivision of Intramural Division Program, National Institute of Mental Health, National Institute of Health, Bethesda, Md., USA; bKey Laboratory for Applied Statistics of Ministry of Education and School of Mathematics and Statistics, Northeast Normal University, Changchun, China Corresponding Author
Yin Yao Shugart
Division of Intramural Research Program, National Institute of Mental Health National Institute of Health, Building 35, Room 3A 1000, 35 Convent Drive
Bethesda, MD 20892 (USA)
Tel. +1 301 496 4341, E-Mail email@example.com
With the advent of sequencing technology opening up a new era of personal genome sequencing, huge amounts of rare variant data have suddenly become available to researchers seeking genetic variants related to human complex disorders. There is an urgent need for the development of novel statistical methods to analyze rare variants in a statistically powerful manner. While a number of statistical tests have already been developed to analyze collapsed rare variants identified by association tests in case-control studies, to date, only two FBAT tests-for-rare (described in the updated FBAT version v2.0.4) have applied collapsing methods analogously in family-based designs. For further research in this area, this study aims to introduce three new beta-determined weight tests for detecting rare variants for quantitative traits in nuclear families. In addition to evaluating the performance of these new methods, it also evaluates that of the two FBAT tests-for-rare, using extensive simulations of situations with and without linkage disequilibrium. Results from these simulations suggest that the four tests using beta-determined weights outperform the two collapsing methods used in FBAT (-v0 and -v1). In addition, both the linear combination method (detailed in the FBAT menu v2.0.4) and the multiple regression method (mixing LASSO and Ridge penalties) performed better than the other two beta-determined weight tests we proposed. Following testing and evaluation, we submitted four new beta-determined weight methods of statistical analysis in a computer program to the Comprehensive R Archive Network (CRAN) for general use.
© 2012 S. Karger AG, Basel