RELATIONSHIPS AMONG EARLY EUROPEAN MAIZE INBREDS .2. COMPARISON OF PEDIGREE AND RFLP DATA
RELATIONSHIPS AMONG EARLY EUROPEAN MAIZE INBREDS .2. COMPARISON OF PEDIGREE AND RFLP DATA
复制标题
DOI：
10.2135/cropsci1993.0011183x003300050014x
复制
发表时间：
1993-09-01
期刊：
影响因子：
2.3
通讯作者：
BOPPENMAIER, J
中科院分区：
文献类型：
--
作者：
MESSMER, MM;MELCHINGER, AE;BOPPENMAIER, J
基因型间遗传关系的信息对育种者非常重要。采用家系分析和RFLP分析方法，对来自欧洲的29个玉米（Zea mays L.）自交系（18个燧石系和11个裂谷系）进行了分析。我们的主要目的是比较基于系谱数据的Malecot共祖先(f)与基于188个克隆-酶组合（CECs）的RFLP数据的遗传相似性（GS），以确定其量化近交系间亲缘程度的能力。有87对燧石系（f > 0）和30对凹痕系（r(s) = 0.86）的相关等级f与GS的相关性极显著（P < 0.01）。由于RFLP数据检测到的家系可能存在错误，并经杂种优势数据证实，因此排除了一个凹痕系。基于f对GS的线性回归，共祖分别解释了相关对燧石系和凹痕系GS变异的82%和70%。因此，同祖先和标记方法都非常适合于(i)测量亲缘关系的平均水平和（ii）识别密切相关的系。观察到的单个对系的GS偏离其回归f可以解释为：(i)随机遗传漂变和选择导致亲本基因组贡献不相等，（ii）由于使用的cec对基因组进行采样而导致GS估计值随机变化，（iii）不相关系的GS估计值存在很大差异，以及（iv）实验误差。如果基因组样本中有足够数量的CECs（大约100个），基于rflp的GS估计比f更准确地反映了给定系对的真实遗传相似性。
Information on genetic relationships among genotypes is of great importance to breeders. In this study we analyzed 29 maize (Zea mays L.) inbreds (18 flint and 11 dent lines) from the European germplasm by means of pedigree and RFLP analyses. Our main objective was to compare Malecot's coancestry (f) based on pedigree data with genetic similarity (GS) based on RFLP data of 188 clone-enzyme combinations (CECs) for their ability to quantify the degree of relatedness among inbreds. Rank correlations between f and GS were highly significant (P < 0.01) for 87 related (f > 0) pairs of flint lines (r(s) = 0.71) and for 30 related pairs of dent lines (r(s) = 0.86). One dent line was excluded because of likely errors in pedigree detected by RFLP data and confirmed by heterosis data. Based on linear regression of GS on f, coancestry explained 82 and 70% of the variation in GS for related pairs of flint and dent lines, respectively. Therefore, both the coancestry and marker approach are well-suited to (i) measure the average level of relatedness and (ii) identify closely related lines. Observed deviations of GS from its regression on f for individual pairs of lines can be explained by (i) random genetic drift and selection causing unequal genomic contributions of the parents, (ii) random variation of GS estimates due to sampling of the genome by the CECs used, (iii) wide variation in the GS estimates of unrelated lines, and (iv) experimental errors. Provided the genome is sampled by an adequate number (> 100) of CECs, RFLP-based GS estimates reflect more accurately the true genetic similarity of a given line pair than f.