A model-based approach to the non-unique oligonucleotide probe selection problem

In order to accurately measure the gene expression levels in oligonucleotide microarray experiments, it is crucial to design ldquouniquerdquo, highly specific and sensitive probes for the identification of biological agents such as genes in a sample. It is difficult to design unique probes for very closely related genes, such as the known strains of HIV genes. The ldquonon-uniquerdquo probe selection problem consists of determining a set of probes, not necessarily unique, that can uniquely identify targets while containing a minimal number of probes. In this paper, we describe a simple model-based method to obtain a near minimal non-unique probe set. Preliminary experimental results are encouraging and at least comparable, if not better, to those obtained by two recently published methods.