应用区间作图法(IM)、复合区间作图法(CIM)、基于混合线性模型的复合区间作图法(MCIM)和多重区间作图法(MIM)对水稻杂交组合（培矮64S×日本晴）F2群体株高性状进行了QTL作图分析。结果表明，（1）在同一显著水平下(α=0.0023)，4种方法共发现10个QTLs，其中IM法7个，CIM法10个，MCIM法3个，MIM法1个。CIM法的发现能力最强。（2）CIM法发现了IM法发现的全部QTLs，和左端标记的距离也基本一致。（3）4种方法检测到了同一QTL，且共同检测到QTL的贡献率最大。（4）4种方法估计的QTL的平均加性效应ā和显性效应 无显著性差异，同一QTL在显性效应的方向上表现一致。（5）建议使用多种作图方法共同作图，并优先标定共同发现的QTL。
Four mapping methods, the interval mapping(IM) , the composite interval mapping(CIM), the mixed composite interval mapping(MCIM) and the multiple interval mapping(MIM), were employed to detect the quantitative trait loci（QTL） of rice plant height with an F2 data from the intercross PA64S×Nipponbare. The results are as follows: (1) 7 significant QTLs were detected by IM，10 QTLs by CIM，3 QTLs by MCIM，1 QTL by MIM for the trait at the same significant level. So CIM had the highest detective capability among the four methods. (2) The loci detected by IM were also detected by the method of CIM, and aiming at the same QTL, their distance to the left marker was almost coincidence.（3）Only a same QTL could be detected by all the four methods, respectively, and the locus had the biggest contribution to plant height. (4) The differences for average of additive effects (absolute values) or dominant effects(practical values) of QTLs were insignificant among four methods and the dominant effects of the same QTLs had the same direction. (5) The epistatic effects could be estimated by MCIM and MIM in the four mapping methods. Only MCIM was employed to estimate the epistatic effects under the significant level at 0.005 because only one QTL was detected by MIM in the paper. (6) A suggestion for mapping QTLs is that multiple methods should be considered, and it is preferential to declare the QTLs confirm simultaneously by the four methods.
全 文 ：