Biplot Analysis of Genotype × Environment Interaction: Proceed with Caution
Biplot Analysis of Genotype × Environment Interaction: Proceed with Caution
复制标题
基因型 × 环境相互作用的双图分析：谨慎进行
DOI：
10.2135/cropsci2008.11.0665
复制
发表时间：
2009
期刊：
影响因子：
2.3
通讯作者：
J. Burgueño
中科院分区：
文献类型：
--
作者：
Rong‐Cai Yang;J. Crossa;P. Cornelius;J. Burgueño
双标图分析已被用于研究基因型×环境互作（GE）或任何双向表。它的描述和可视化能力沿着用户友好的软件，使植物科学家能够通过点击计算机按钮来检查任何双向数据。尽管被广泛使用，双标图分析的有效性和局限性还没有被完全检验。在这里，我们确定并详细讨论了围绕双标图分析的过度使用或滥用的六个关键问题。我们的问题是：（i）双标图分析中前两个乘法项的保留是否足够;（ii）双标图是否不仅仅是一种简单的描述性技术;（iii）从双标图中识别“哪个赢在哪里”模式的现实程度如何;（iv）如果基因型和/或环境是随机效应呢？（v）双标图分析对理解相互作用的性质和原因有多大的相关性;以及（vi）双标图分析对交叉相互作用的检测有多大的贡献。我们强调需要使用双标图中个体基因型和环境得分的置信区域，以基于统计检验对基因型选择或品种推荐做出关键决策。我们的结论是，双标图分析只是一个直观的描述性统计工具，研究人员应该谨慎进行，如果使用双标图分析超出这个简单的功能。
Biplot analysis has been used for studying genotype × environment interaction (GE) or any two-way table. Its descriptive and visualization capabilities along with the availability of userfriendly software have enabled plant scientists to examine any two-way data by a click on a computer button. Despite widespread use, the validity and limitations of biplot analysis have not been completely examined. Here we identify and briefl y discuss six key issues surrounding overutilization or abuse of biplot analysis. We question (i) whether the retention of the fi rst two multiplicative terms in the biplot analyses is adequate; (ii) whether the biplot can be more than a simple descriptive technique; (iii) how realistic a “which-won-where” pattern is identifi ed from a biplot; (iv) what if genotypes and/ or environments are random effects; (v) how relevant biplot analysis is to the understanding of the nature and causes of interaction; and (vi) how much the biplot analysis can contribute to detection of crossover interaction. We stress the need for use of confi dence regions for individual genotype and environment scores in biplots to make critical decisions on genotype selection or cultivar recommendation based on a statistical test. We conclude that the biplot analysis is simply a visually descriptive statistical tool and researchers should proceed with caution if using biplot analysis beyond this simple function.