Content based searching of gene expression databases using binary fingerprints of differential expression profiles

Microarray technologies are commonly used for large scale measurement of gene expression. Microarray studies reveal information between different biological processes and produce hypotheses for further investigation. Availability and rapid growth of microarray databases has made integrated analysis of these databases computationally challenging. We present a new approach to content based searching in microarray databases, using a binary fingerprint approach that is inspired from the chemoinformatics field. Differential expression profiles are represented in concise bit vectors to allow rapid comparison of microarray experiments using bit-wise operations. We demonstrate that the fingerprint representation provides fast search times, while preserving accuracy when compared to the more commonly used correlation-based metrics.