LASSO算法在基于自然群体随机交配的多iQTL定位中的应用

Multi-iQTL mapping using LASSO algorithm based on random mating design in natural population

基于自然群体随机交配,应用LASSO算法进行多iQTL定位,带有坐标下降的LASSO方法能够将无关变量压缩并使之趋于零.对模拟产生的自然群体和水稻自然群体数据进行模拟研究,结果表明应用LASSO算法进行多iQTL定位,能够高效、准确地定位到iQTL的位置,并且对各种遗传效应能进行较精确的估计.

For random mating design in natural population, LASSO algorithm with coordinate descent method can compress independent vector to zero when mapping multiple QTLs. In this study, stimulated and natural rice population were both analyzed. The results proved that LASSO can not only efficiently locate iQTLs but also estimate corresponding genetic effects accurately.