Comparing performances of different statistical models and multiple threshold methods in a nested association mapping population of wheat.

Autor:	Sandhu KS; Department of Crop and Soil Sciences, Washington State University, Pullman, WA, United States., Burke AB; Department of Crop and Soil Sciences, Washington State University, Pullman, WA, United States., Merrick LF; Department of Crop and Soil Sciences, Washington State University, Pullman, WA, United States., Pumphrey MO; Department of Crop and Soil Sciences, Washington State University, Pullman, WA, United States., Carter AH; Department of Crop and Soil Sciences, Washington State University, Pullman, WA, United States.
Jazyk:	angličtina
Zdroj:	Frontiers in plant science [Front Plant Sci] 2024 Oct 01; Vol. 15, pp. 1460353. Date of Electronic Publication: 2024 Oct 01 (Print Publication: 2024).
DOI:	10.3389/fpls.2024.1460353
Abstrakt:	Nested association mapping (NAM) populations emerged as a multi-parental strategy that combines the high statistical power of biparental linkage mapping with greater allelic richness of association mapping. Several statistical models have been developed for marker-trait associations (MTAs) in genome-wide association studies (GWAS), which ranges from simple to increasingly complex models. These statistical models vary in their performance for detecting real association with the avoidance of false positives and false negatives. Furthermore, significant threshold methods play an equally important role for controlling spurious associations. In this study, we compared the performance of seven different statistical models ranging from single to multi-locus models on eight different simulated traits with varied genetic architecture for a NAM population of spring wheat ( Triticum aestivum L.). The best identified model was further used to identify MTAs for 11 different agronomic and spectral reflectance traits, which were collected on the NAM population between 2014 and 2016. The "Bayesian information and linkage disequilibrium iteratively nested keyway (BLINK)" model performed better than all other models observed based on QQ plots and detection of real association in a simulated data set. The results from model comparison suggest that BLINK controls both false positives and false negatives under the different genetic architecture of simulated traits. Comparison of multiple significant threshold methods suggests that Bonferroni correction performed superior for controlling false positives and false negatives and complements the performance of GWAS models. BLINK identified 45 MTAs using Bonferroni correction of 0.05 for 11 different phenotypic traits in the NAM population. This study helps identify the best statistical model and significant threshold method for performing association analysis in subsequent NAM population studies. Competing Interests: The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer PM and the reviewer DKS declared a past co-authorship with the author KS to the handling editor. (Copyright © 2024 Sandhu, Burke, Merrick, Pumphrey and Carter.)
Databáze:	MEDLINE
Externí odkaz:	Zobrazit plný text záznamu