An interval mappingprocedure based on the random model approach was applied to investigate itsrobustness and properties for QTL mapping in populations with prevailing half-sibstructures. Under random model, QTL location and variance components were estimatedusing maximum likelihood procedures. The estimation of parameters was basedon sib-pair approach. The proportions of genes identical-by-descent (IBD) atthe two QTLs were estimated from the IBD at two flanking marker loci. Estimatesfor QTLs parameters (locations and variance components) and power of detection,in a trait with h2 = 0.25, were obtained using simulated data, varyingthe number of families, number of half-sibs by families, proportion of QTL varianceand QTLs positions. These QTLs were positioned at the same interval, at adjacentintervals and at no adjacent intervals. The most important factors influencingthe estimates of QTLs parameters and power were the proportion of variance dueto QTLs the number and size of the half-sibs families. The most biased variancecomponents estimations were obtained when sample size was ten. Given a sufficientnumber of families and high proportions of genetic variance due to QTLs, therandom model approach can detect a QTLs with high power and provides accurateestimates of the QTLs position if two QTLs are not present in the same interval.For fine QTL mapping and proper estimation of QTLs variance, more sophisticatedmethods are required.
