Cross‐platform comparisons of algorithms for calculating real‐time PCR amplification efficiencies

Abstract

The analysis of real‐time PCR data has become the focus of mathematical modeling to increase quantification precision and accuracy. Methods include: (1) the slope‐derived efficiency calculation from the standard curve method using Ct values determined either from (a) the fit‐point method or (b) the second derivative maximum (SDM) of a four parametric logistic model; or (2) the single amplification plot methods to compute efficiencies from individual PCR kinetic curves using comprehensive algorithms such as (a) mid‐point regression (Data Analysis for Real‐time PCR or DART‐PCR), (b) the window‐of‐linearity algorithm (LinReg) or (c) noise‐resistant iterative nonlinear regression (Real‐time PCR Miner). We directly compared these methods on three commonly used real‐time PCR instruments (BioRad iCycler iQ™, Stratagene Mx3005P™ and ABI 7500 Fast PCR System). SYBR‐Green® based RT2 Profiler™PCR Arrays from SuperArray Bioscience were used with serially diluted DNA templates and the amplification efficiency for each PCR assay was computed. This study demonstrates that Real‐time PCR Miner provides the best precision in efficiency estimation independent of the PCR instrument, while the precisions for other methods are platform‐dependent. Accounting for amplification efficiencies in fold‐change calculations may increase the precision of gene expression measurements using real time PCR.