Advances in digital microfluidic (DMF) technologies offer a promising platform for a variety of biochemical applications, ranging from massively parallel DNA analysis and computational drug discovery to toxicity monitoring and medical diagnosis. In this paper, we address the migration problem that arises when the technology undergoes a change in the context of DMFs. Given a biochemical reaction synthesized for actuation on a given DMF architecture, we discuss how the same biochemical reaction can be ported seamlessly to an enhanced architecture, with possible modifications to the architectural parameters (e.g., clock frequency, mixer size, and mixing time) or geometric changes (e.g., change in reservoir locations or mixer positions, inclusion of new sensors or other physical resources). Complete resynthesis of the protocol for the new architecture may often become either inefficient or even infeasible due to scalability, proprietary, security, or cost issues. We propose an adaptation method for handling such technology-changes by modifying the existing actuation sequence through an incremental procedure. The foundation of our method lies in symbolic encoding and satisfiability-solvers, enriched with pertinent graph-theoretic and geometric techniques. This enables us to generate functionally correct solutions for the new target architecture without necessitating a complete resynthesis step, thereby enabling the utilization of these chips by users in biology who are not familiar with the on-chip synthesis tool-flow. We highlight the benefits of the proposed approach through extensive simulations on assay benchmarks.