UC Riverside

The motivation of our work is to demonstrate that scientists can describe biochemical assays using a high-level domain-specific programming language and can execute them automatically on a two-dimensional electrowetting array featuring integrated sensors that provide feedback in real-time to the host PC that controls the integrated system. The fundamental research problem that this paper addresses is how to express, at a high level of abstraction, the acquisition of sensory information from the device, and the computation on that data, which is required to perform real-time decision making in response. The approach that we have taken is first, to create a new version of BioCoder, a programming language for automated biology, which we have specialized for electrowetting arrays featuring integrated sensors, and second, to use this language to specify two feedback-driven Polymerase Chain Reaction (PCR) assays at the desired level of abstraction. Our methodology is to evaluate the performance of these assays using a custom-build runtime system to control the execution of feedback-driven assays on a cycle-accurate software simulator that accurately characterizes the behavior of the electrowetting platform during assay execution. The result of this experiment is successful simulation of these non-trivial feedback-driven assays, starting from a high-level specification, which allows us to conclude that high-level programming language design and implementation targeting electrowetting (or other competing laboratory-on-a-chip technologies) is feasible.