Investigation And Simulation Studies On Factors Influencing QTL Mapping

The objective of quantitative trait loci(QTL)linkage mapping is to obtain the genetic information and mechanism of a quantitative trait,which is the prerequisite for choosing appropriate breeding methods and gene cloning in practice.The factors influencing the accuracy of QTL mapping include the type of mapping population,the number(density)of molecular markers,the sample size of mapping population,the heritability of the trait and the QTL,the genetic difference between parental lines,the statistical method of linkage mapping.It is important to carefully choose these factors before QTL mapping.1.In the present study,the QTL data were collected from real QTL studies published on Theoretical and Applied Genetics(TAG)in recent 18 years,and the factors influencing the accuracy of QTL mapping were analyzed.We mainly discussed the influence of the heritability of traits,the sample size of mapping population,the type of mapping population,the number of genetic markers and statistical method on QTL detection.The results are as follows:(1)Descriptive statistical analysis showed that the most widely used mapping population was RIL population,the most commonly used method for linkage mapping is CIM.There is an increasing tendency of the sample size and the number of genetic markers.There is a significant impact of mapping population type on the number of QTL detection,DH>RIL>BC>F2 for the average number of detected QTL,and the number of QTL detected was 3.3,3.0,2.8 and 2.6,respectively.(2)Poisson regression analysis showed that three was an extremely significant association between the sample size of mapping population and the number of QTL detection,the type of mapping population and different method of linkage mapping,while the influence of the number of molecular markers on QTL detection is also significant.For population type,more QTL were detected by using RIL and DH population than BC or F2.The influence(in term of regression coefficient)of trait heritability on QTL detection was greater than sample size and the number of markers.For the statistical method of linkage mapping,ICIM,MCIM and MIM can detect more QTL than CIM or IM,while CIM can detect more QTL than IM.(3)The distribution of the percentage of phenotypic variance explained by QTL showed that the genetic system of quantitative trait in plant is mainly major gene plus polygene mixed inheritance,while major inheritance model is rare.As it is found that the most used mapping population is RIL population,the most commonly used method for linkage mapping is CIM through analysis of the instance data,RIL population with the CIM method was used for simulation study.The influence of several factors on the QTL mapping by setting up different levels of trait heritability,the number of markers(density),sample size and the number of the QTL controlling quantitative trait were analyzed.The exponential distribution and normal distribution were used for QTL effect simulation.The false discovery rate,power for QTL detection and the percentage of phenotypic variance explained by the QTL(PVE)were evaluated.The results were summarized as follows:(1)If the number of the QTL,the sample size and the number of markers were fixed,the detection power of QTL and PVE increase along with the increase of the trait heritability,indicating that improving trait heritability is the most effective way to improve PVE.(2)If the number of the QTL,the sample size and the trait heritability were fixed,the PVE and detection power increase along with the increase of the marker number,but the range of PVE is relatively small.When the density of markers was 5cM,its influence on QTL detection power and PVE are very small with the increase of the marker density,indicating that the marker density is saturated.(3)If the number of the QTL,marker density and the trait heritability were fixed,the PVE and the detection power increase along with the increase of the sample size,the detection power of QTL increase much more than PVE,indicating that improving sample size is the most effective way to improve the detection power,but when the QTL number is 10,20,50,the false discovery rate also increase slightly with the increase of sample size.When the sample size was 3000,its influence on detection power and PVE are very small along with the increase of marker density,indicating that the sample size is saturated.(4)If the sample size,marker density and the trait heritability were fixed,the detection power of QTL and PVE decrease along with the increase of the number of QTL,but the false discovery rate increase firstly and then began to decrease.When the number of QTL was 100,the detection power and PVE were relatively stable along with the increase of the number of the QTL.(5)If the sample size,marker density,trait heritability and the number of the QTL were fixed,the QTL detection power of exponentially distributed QTL effect is higher than that of normally distributed QTL effect.