Plant & Animal Genomes XIII Conference

January 15-19, 2005
Town & Country Convention Center
San Diego, CA

Fine QTL Mapping Using Combined Multipletrait-Multilocus Or Multipleenvironment-Multilocus Analyses

Ori Orion , Yefim I. Ronin , Dina G. Minkov , Abraham B. Korol

  University of Haifa, Evolution Institute, Mount Carmel, Haifa 31905 Israel

We describe here a practical high resolution approach for QTL analysis allowing extraction maximum mapping information contained in the data. Using MultiQTL package, we demonstrate, for both simulated and experimental data sets, that combined MultipleTrait-Multilocus (MLT-MIM) analysis allows to dramatically improve several characteristics of the mapping quality compared with the standard single-trait chromosome-wise analysis without increase in sample size. This includes increase of QTL detection power, reduced bias of the estimated QTL position and the size of its confidence interval (for one of the simulated examples, from 15 cM for the single-trait single-chromosome model, 11.6 cM for single-chromosome-MLT, 5.8 cM for single-trait-MIM, and 1.9 cM for MLT-MIM), as well as the precision of the estimated QTL effect. It is noteworthy that single-trait-MIM may result in an increased level of false positive although transition to MLT-MIM may significantly improve the situation. Similar results were obtained with simulated datasets for situations when a target trait is scored in multiple environments. Namely, joint analysis of the trait scores across environments conducted using multiple-interval mapping model allows increasing QTL detection power, narrowing the confidence interval of estimated QTL position, and higher power of detection of QTL-environmental interaction. All the foregoing effects were also found on real experimental data.