B-040 Hyperspectral Detection and Unmixing of Multiple Bound Fluorophores for qPCR

Background Multiplexed bead-based assays are valuable tools in laboratory medicine, allowing simultaneous measurement of multiple analytes. However, spectral overlap between fluorophores can complicate accurate quantification. We developed and demonstrate a custom plate reader system incorporating a Hamamatsu C12880MA mini-spectrometer with modified principal component analysis (PCA) capabilities to address these challenges and evaluate its performance with four-color fluorescent beads. Methods A custom plate reader constructed using a Hamamatsu C12880MA mini-spectrometer is validated using TetraSpeck™ microspheres (0.5 µm, fluorescent blue/green/orange/dark red) dispensed into a Cellvis 96-well plate. Principal component decomposition is performed on the spectral data to separate and quantify the individual fluorescent components. The system*s ability to distinguish between different colored beads and accurately decompose overlapping spectra was assessed. Results Initial testing demonstrates the plate reader*s capability to detect and measure fluorescent signals from the TetraSpeck™ microsphere samples across all four spectral ranges (blue/green/orange/dark red). Principal component analysis successfully decomposes the spectral data, allowing for distinction between different bead populations. While the dataset is limited, preliminary results suggest the system can effectively separate overlapping spectral signatures. Conclusion This proof-of-concept study demonstrates the utility of our custom plate reader system, built around a low cost, commercial OTS mini-spectrometer with modified PCA for multi-color bead analysis. Further validation with larger sample sizes and different bead combinations is necessary to fully characterize the system*s performance and establish its practical applications in laboratory testing. While populations of beads separated entirely, further development is required for quantitative analysis of fluorophore concentrations. We are extending the study to time- resolved fluorescence in lanthanide tags. Further development of this technique will use support- vector machines for classification of clusters, extending analysis to ten-color systems.