Composite Interval Mapping Of Quantitative Trait Location And Software Development In A Full-sib Family Of Forest Trees

Because of their complex biological characteristics, such as high heterozygosity and long lifecycle, it is almost impossible to generate inbred lines in forest trees. As a result, the statisticalmodels or softwares for mapping quantitative trait location(QTL) in inbred lines cannot bedirectly applied to forest trees. This leads to a slow advancement in forest molecular quantitativegenetics. Therefore, it is necessary to develop statistical models and computer softwares formapping QTLs by incorporating the unique characteristics of forest trees.Based on the genetic linkage map constructed with various segregation types of molecularmarkers, we established a statistical model to implement the method of composite interval QTLmapping in a forest full-sib family, which can be produced by crossing two parents with somephenotypic traits substantially different. In the mean time, the corresponding computer programwas developed according to the procedures of composite interval mapping, which includedseveral C subfunctions like QtlCimMap(), opendata(), GenerateMarkType(), FreqRecomABQ(),and Single(). The linkage-group-wide threshold of0.05significant level, for asserting whether ahigh peak of LOD profile is a QTL or not, was determined by permutation tests. The statisticalmodel and the computer program were evaluated or tested through Monte Carlo simulations.Finally, we applied the novel QTL mapping method to a real dada in Populus and identified aQTL that controlled the rooting ability at the position of13cM in the11thinterval on linkage3.The study provided a powerful analytical tool for accurately detecting QTLs in forest trees.