CIM Confidence Intervals

[Nathaniel]

Hi Karl,

When calculating confidence intervals, typically I have been advised to use the scanone output as opposed to the CIM. The rationale is that the CIM may give a false understanding of the true length of the QTL. I can see how the interval would depend on the # of covariates and window size used when calculating the CIM, and in the guidebook it is advised to use the CIM for exploratory purposes as opposed to "general use". What would be the more proper justification for using the scanone to calculate confidence intervals instead of the CIM?

Thanks,

Nathaniel