Pattern-mapped Multiple Detection of 11 Pathogenic Bacteria Using a 16s RDNA-based Oligonucleotide Microarray

Pathogen detection is an important issue in human health due to the threats posed by severe communicable diseases. In the present study, to achieve efficient and accurate multiple detection of 11 selected pathogenic bacteria, we constructed a 16S rDNA oligonucleotide microarray containing doubly specific capture probes. Many target pathogens were specifically detected by the microarray with the aid of traditional perfect match-based analysis using our previously proposed two-dimensional visualization plot tool. However, some target species or subtypes were difficult to discriminate by perfect match analysis due to nonspecific binding of conserved 16S rDNA-derived capture probes with high sequence similarity. We noticed that the patterns of specific spots for each strain were somewhat different in the two-dimensional gradation plot. Therefore, to discriminate subtle differences between phylogenetically related pathogens, a pattern-mapping statistical model was established using an artificial neural network algorithm trained by experimental repeats. The oligonucleotide microarray system harboring doubly specific capture probes combined with the pattern-mapping analysis tool resulted in successful detection of all target pathogens including even subtypes of two closely related species showing strong nonspecific binding. Collectively, the results indicate that our novel combined system of a 16S rDNA-based DNA microarray and a pattern-mapping statistical analysis tool is a simple and effective method for detecting multiple pathogens.