Multiplexed Digital Assay Data Exclusion for Target Concentration

The patent segments the partition populations into distinct groups based on their target content profiles. By clustering partitions with similar amplification patterns together, the method separates overlapping signals from different target combinations, enabling accurate identification and quantification of individual targets even in multiplexed assays where multiple targets coexist in the same partition.

The patent transforms the analysis from a single-dimensional signal amplitude measurement to a multi-dimensional space by considering amplification patterns across multiple targets and partitions simultaneously. This dimensional expansion allows the resolution of overlapping signals through pattern recognition and clustering algorithms that identify distinct population groups based on their unique target content signatures.

The patent extracts and isolates specific partition populations that contain unambiguous signals for individual targets by excluding partitions where signal overlap occurs. By identifying and removing contaminated data points from the calculation set, the method maintains statistical reliability from sufficient valid partitions while eliminating the precision-deteriorating effect of overlapping populations.

The patent applies partial action by using only the necessary subset of partition data that provides reliable concentration information, rather than forcing the use of all available partitions. This selective data inclusion ensures that concentration calculations are based on clear, unambiguous signals while still maintaining adequate statistical power from the remaining valid partitions.