ABSTRACT Infections by different pathogens can manifest with similar symptoms, but appropriate treatment requires specific and accurate diagnosis. Clinicians often turn to multiplexed assays testing for many organisms (e.g. BioFire). While these approaches can test for 50-70 organisms, they do not provide concentration titers, which is necessary to identify the causative pathogen among the several false positives or clinically meaningless commensals. As a result, the clinician must perform additional tests to identify which of the positives is causative. Although these tests use quantitative PCR, in clinical labs the results are reported as presence/absence due to the finicky nature of PCR in this setting, which is sensitive to minor variations in reaction efficiency, operator variability. As a result, today, only a few widespread PCR tests are FDA approved to report quantitative result. In contrast to qPCR, digital PCR (dPCR) measures target titers by counting individual molecules. As a result, dPCR provides an absolute concentration measurement that doesn’t require a standard curve. In addition, the reaction is cycled to endpoint, then quantified; it does not require careful estimation of the amplification rate, which is a major source of variability in qPCR. Thus, dPCR is less sensitive to variations in reaction efficiency and provides superior consistency. However, current dPCR methods are limited in multiplexing, allowing just 5- 6 targets per assay, while qPCR can test up to 100. Moreover, dPCR requires complex microfluidic equipment that burdens testing lab personnel and increases cost. Until these issues can be addressed, qPCR will continue to dominate the clinical lab, and quantitative and absolute pathogen load reporting will remain beyond reach. Here, we propose a novel nucleic acid technology combining the quantitativeness and robustness of dPCR with the simplicity and multiplexing of qPCR. Our vision is to enable broad spectrum detection wherein each pathogen is associated with a high confidence, quantitative titer. Our approach – gigapixel PCR (gPCR) – is enabled by our recent discoveries of self-assembled partitioning, for microfluidic-free generation of monodispersed emulsions, and linearized target quantitation with capillary electrophoresis (CE). CE allows sensitive quantitation over 7 decades and provides amplicon length information with single nucleotide resolution. In gPCR, we use this to perform multiplexed detection of over 100 amplicons per reaction. In contrast to qPCR, which requires that the sample be split to test for different targets, thereby diluting it and reducing sensitivity, with gPCR the targets are tested without splitting, maintaining them at maximal concentration, and substantially increasing sensitivity. Moreover, based on robust dPCR, gPCR provides reproducible, quantitative results across testing conditions. It thus addresses the major limitations of current dPCR technologies and provides the first viable alternative to qPCR in the clinic. We will develop and validate the technology against accepted standards (SeraCare), and work with our longstanding collaborators (Drs. Melanie Ott and Charles Chiu) to apply it to respiratory and CNS infections from samples previously collected at UCSF hospitals.

NARRATIVE Hospital infections must be diagnosed and treated rapidly to minimize spread and maximize patient outcome. This proposal will deliver a novel infectious disease diagnostic that will arm clinicians with vital quantitative information across many candidate pathogens, allowing them to diagnose and treat infections more quickly and accurately.