Robust regression methods for real-time polymerase chain reaction

Abstract

Current real-time polymerase chain reaction (PCR) data analysis methods implement linear least squares regression methods for primer efficiency estimation based on standard curve dilution series. This method is sensitive to outliers that distort the outcome and are often ignored or removed by the end user. Here, robust regression methods are shown to provide a reliable alternative because they are less affected by outliers and often result in more precise primer efficiency estimators than the linear least squares method.

Keywords: Outliers; PCR efficiency estimation; Real-time PCR; Robust regression; Standard curve; qPCR.