An Adaptive Sequential Replacement Method for Variable Selection in Linear Regression Analysis  

作　　者：Jixiang Wu Johnie N. Jenkins Jack C. McCarty Jr. Jixiang Wu;Johnie N. Jenkins;Jack C. McCarty Jr.(Genetics and Sustainable Agriculture Research Unit, USDA-ARS, Mississippi State, USA)

机构地区：[1]Genetics and Sustainable Agriculture Research Unit, USDA-ARS, Mississippi State, USA

出　　处：《Open Journal of Statistics》2023年第5期746-760,共15页统计学期刊（英文）

摘　　要：With the rapid development of DNA technologies, high throughput genomic data have become a powerful leverage to locate desirable genetic loci associated with traits of importance in various crop species. However, current genetic association mapping analyses are focused on identifying individual QTLs. This study aimed to identify a set of QTLs or genetic markers, which can capture genetic variability for marker-assisted selection. Selecting a set with k loci that can maximize genetic variation out of high throughput genomic data is a challenging issue. In this study, we proposed an adaptive sequential replacement (ASR) method, which is considered a variant of the sequential replacement (SR) method. Through Monte Carlo simulation and comparing with four other selection methods: exhaustive, SR method, forward, and backward methods we found that the ASR method sustains consistent and repeatable results comparable to the exhaustive method with much reduced computational intensity.With the rapid development of DNA technologies, high throughput genomic data have become a powerful leverage to locate desirable genetic loci associated with traits of importance in various crop species. However, current genetic association mapping analyses are focused on identifying individual QTLs. This study aimed to identify a set of QTLs or genetic markers, which can capture genetic variability for marker-assisted selection. Selecting a set with k loci that can maximize genetic variation out of high throughput genomic data is a challenging issue. In this study, we proposed an adaptive sequential replacement (ASR) method, which is considered a variant of the sequential replacement (SR) method. Through Monte Carlo simulation and comparing with four other selection methods: exhaustive, SR method, forward, and backward methods we found that the ASR method sustains consistent and repeatable results comparable to the exhaustive method with much reduced computational intensity.

关 键 词：Adaptive Sequential Replacement Association Mapping Exhaustive Method Global Optimal Solution Sequential Replacement Variable Selection 

分 类 号：TN9[电子电信—信息与通信工程]