I am currently building a threshold model that partitions observations into those above and below the level i. i.e.
model_threg <- pgmm(ML ~ ...+ I(A*I(B>=