The concept was apparently introduced in ^[1] and is given by the full wave function $\Psi$ of a system with some coordinates fixed at the position of the Bohmian trajectories for some fixed initial position. Thus it is possible to define the notion of the wave function of a subsystem with coordinates $x$, with the rest of the system described in coordinates $y$. If we assume the coordinates $y$ following the Bohmian trajectory $Y(t)$ in time then the conditional wave function of the subsystem is \[ \psi(t,x) = \Psi(t,x,Y(t)). \]

Of course we cannot expect this quantity to follow a linear evolution equation like the full wave function. Also it will not stay normalized with the passage of time and even includes the wave function collapse naturally. Here is a derivation of the conditional evolution equation.