Hardware Implementation of Denoising Algorithms for Nanopore Sensing

Effective biosensors continue to be a research area of great interest for both defense and medical applications. In particular, silicon pores with diameters in the range of micro/nano-meters have demonstrated the ability to detect an array of analytes. Typically, these sensors make use of the Coulter counter set-up where a drop in current across the chamber is observed when a biomaterial passes. The duration and amplitude of this drop is indicative of the biomaterial's size and shape.In order to effectively use such sensors, however, robust denoising and classification algorithms must also be developed. Recently, Non-Positive Go Decomposition (NpGoDec) was shown to be an effective denoising method for biological data, correctly classifying simulated controlled data for Immunoglobulin G biomolecule with 96 percent accuracy.In this work, a research team for Arizona State University programmed the NpGoDec algorithm onto a Field Programmable Gate Array (FPGA) for on-chip, biosensor processing. There are several benefits to such a system. First, denoising the signal on an FPGA reduces processing time by avoiding the transmission of the raw data into off-line processing software such as MATLAB and brings biological sensing one step closer to real time. In addition, performing much of the signal processing work on the FPGA moves the sensor closer to being a portable device. The system is carefully investigated for accuracy and processing time as compared to the original, simulated signal. Our approach also enables the integration of a classifier onto an FPGA, which will allow the system to quickly identify the biomaterials passing through a nanopore.