In genetic studies, many genetic function-valued traits have temporal or spatial structures. The current methods for mapping these traits require specific parameters and they are difficult to get or impractical in many scenarios. To solve the problem, the authors proposed a general functional regression approach based on estimating equations that is robust. It provides a good compromise between model simplicity, statistical efﬁciency, and computational speed. They tested the new method in circadian mouse behavioral data in 89 N1 backcross mice between C57BL/6J and 129S1/SvImJ strains. For detailed protocols see Xiong et al 2011.