Target estimation in real-time polymerase chain reaction using sequential Monte Carlo

Polymerase Chain Reaction (PCR) is a technique which relies on a polymerase enzyme to amplify a short DNA sequence (target analyte) by replicating it through a series of thermal cycles. Determining the initial number of target analytes is often the goal of quantitative PCR (QPCR). QPCR is used in several applications including detection and quantification of nucleic acids, viral load detection, and gene expression studies, to name just a few. The number of analytes generated at the end of each PCR cycle can be modeled by a Galton-Watson branching process. In this paper, we employ the sequential Monte Carlo technique (a particle filter) to jointly estimate the number of target analytes at the beginning of the QPCR reaction and the efficiency of the reaction. Simulation and experimental results demonstrate that the proposed approach outperforms competing techniques.