A compressive sensing microarray (CSM) is a new device for DNA-based identification of target organisms that leverages the nascent theory of compressive sensing (CS). In contrast to a conventional DNA microarray, in which each genetic sensor spot is designed to respond to a single target organism, in a CSM each sensor spot responds to a group of targets. As a result, significantly fewer total sensor spots are required. In this paper, we study how to design group identifier probes that simultaneously account for both the constraints from the CS theory and the biochemistry of probe-target DNA hybridization. We employ belief propagation as a CS recovery method to estimate target concentrations from the microarray intensities.
Published in:
Computational Advances in Multi-Sensor Adaptive Processing, 2007. CAMPSAP 2007. 2nd IEEE International Workshop on
Date of Conference: 12-14 Dec. 2007