U.S. Genomics is applying single-molecule DNA mapping technology to solve many of the limitations of current diagnostic tests for infectious disease. Existing methods typically require either lengthy culturing steps or rely on PCR amplification approaches that offer limited multiplexing. In contrast, U.S. Genomics' approach aims to use a single reagent set for rapid detection of a wide range of infectious organisms, from a variety of sample types. Reagents and assays are being developed in the following areas:

• Hospital Acquired Infections (HAI)
• Strain-typing & epidemiology
• Environmental monitoring
• Bio-pharmaceutical process monitoring
• Food safety

Approach
U.S. Genomics's genomic mapping technology (hyperlink) consists of engineered systems, optimized reagents, and data analysis tools that can be tailored to a variety of infectious disease diagnostics applications. Under the U.S. Genomic's approach DNA is recovered from microorganisms in a fully automated system that efficiently preserves DNA integrity. The DNA is then prepared and tagged for single-molecule detection by using signal-generating (SG) pairs, which consist of estriction enzymes (RE) and fluorophore-labeled probes. The RE is chosen to yield DNA fragments of a size range that provides optimal stretching efficiency and information content. Probes are designed to recognize short sequences (e.g., 6-8 base-pairs) along a stretched DNA molecule. Serial measurement of the tags, and the separation distance between tags on individual molecules leads to the creation of unique signals, or barcodes, that can be used to accurately identify the host organism for each DNA fragment. In order to generate barcodes it is necessary to first recover high-quality long fragments of genomic DNA. U.S. Genomics has developed methods and engineered components to execute a number of critical steps in-line with the single-molecule detection to provide for optimal results. The integrity of DNA recovered from the automated process has been verified in off-line pulsed-field

gel-electrophoresis (PFGE) experiments (Figure A). Real-time data collection from every molecule is used to generate plots of the DNA length versus the backbone intensity, referred to as comet plots (Figure B, upper). The accumulation of molecules of similar size in the comet plots yields size-based information in a fraction of the time of conventional PFGE. However, more importantly, within each size band of the comet plot, DNA fragments of similar size from many different organisms can be readily discriminated from one another based on their independent and unique barcodes (Figure B, lower)

A data set is analyzed by comparing single molecule barcodes to templates in a data base. The templates for pathogenic organisms that populate the database can be generated either in silico using sequence information, or by using an unsupervised learning algorithm to cluster patterns that result from processing an unsequenced organism.  Data analysis can be performed real-time or offline, and can be archived for subsequent more detailed analysis.
In order to efficiently optimize selection of the SG pair U.S. Genomics has developed an in silico toolset that can predict which elements of the SG pair (restriction enzyme and probes) yield optimum resolution, based on known sequence information. Comparative analysis of individual barcodes from a list of potential organisms (or strain variations) leads to the generation of a score. These scores are accumulated in a heatmap (Figure C) for a list of organisms and a list of probes. Low scores (green) indicate that for a given probe, barcodes can be generated that are unique to the entire list of organisms.