I found no options to control this behavior but if I use Beagle and the target dataset (Simons Genome Diversity Project) is quite different from the reference panel i used (1000 genome), i always have many sites removed, although I have set "impute=false", so what is the cause to this event? and what should I do?