Semi-flexible vs. fully flexible segments question

I am perfoming a docking simulation by Haddock and I want to discuss the differences between semi-flexible and fully flexible segments, and obtain advice for docking a two domain protein with a flexible interdomain linker. I understand the semi-flexible segments to be important in the interface realignment and optimized docking by allowing residues involved in binding to have some degree of flexibility during simulated annelling. In this case, if I were to manually select the interdomain linker segment to be semi-flexible, then all other potential interfaces are set to rigid. Is this correct? And so, is it best to leave the semi-flexible option to automatic and to set to fully flexible the interdomain linker?

Indeed best option is to keep the automatic definition of semi-flexible segments and define the linker as fully flexible

Also you could consider splitting your system into two domains. Check the following paper: