How good is a PCR efficiency estimate: Recommendations for precise and robust qPCR efficiency assessments

Abstract

We have examined the imprecision in the estimation of PCR efficiency by means of standard curves based on strategic experimental design with large number of technical replicates. In particular, how robust this estimation is in terms of a commonly varying factors: the instrument used, the number of technical replicates performed and the effect of the volume transferred throughout the dilution series. We used six different qPCR instruments, we performed 1-16 qPCR replicates per concentration and we tested 2-10 μl volume of analyte transferred, respectively. We find that the estimated PCR efficiency varies significantly across different instruments. Using a Monte Carlo approach, we find the uncertainty in the PCR efficiency estimation may be as large as 42.5% (95% CI) if standard curve with only one qPCR replicate is used in 16 different plates. Based on our investigation we propose recommendations for the precise estimation of PCR efficiency: (1) one robust standard curve with at least 3-4 qPCR replicates at each concentration shall be generated, (2) the efficiency is instrument dependent, but reproducibly stable on one platform, and (3) using a larger volume when constructing serial dilution series reduces sampling error and enables calibration across a wider dynamic range.

Keywords: ANCOVA, analysis of covariance; Amplification efficiency; CLSI, Clinical and Laboratory Standards Institute; Cq, cycle of quantification; Dilution series; E, PCR efficiency; EPA, Environmental protection agency; FDA, food and Drug Administration; GMO, genetically modified organism; IEC, International Electrotechnical Commission; ISO, International Organization for Standardization; MIQE, minimum information for publication of quantitative real-time PCR experiments; NTC, no template control; RIN, RNA Integrity Number; RT-qPCR, reverse transcription-quantitative polymerase chain reaction; Real-time quantitative PCR; Standard curve; qPCR; qPCR assay validation.

Fig. 1

Impact of PCR efficiency on fold change between two conditions. Model shows how determination of efficiency impacts relative fold change calculated in terms of difference in Cq values (ΔCq).

Fig. 2

PCR efficiency estimated for three selected assays using six different instruments. Error bars indicate 95% confidence interval. Efficiencies were estimated from standard curves based on ten-fold dilutions based on six concentrations, each analyzed in four qPCR replicates.

Fig. 3

Effect of transfer volume when constructing standards by means of serial dilution. Four dilution series were constructed using each of the transfer volumes 2 μl, 5 μl and 10 μl. Each standard curve was based on ten-fold dilutions for 6 concentrations, each analyzed in four qPCR replicates. This gave a total of 16 qPCR measurements at each dilution step for each transfer volume tested, and were subject to Monte Carlo analysis. Missing data were not replaced or imputed in the analysis.

Fig. 4

Effect of the transfer volume used in the serial dilution to construct the standards on the number of positive amplification reactions (hits) and on the repeatability of technical replicates at the most diluted standard. Transfer volumes were 10 μl, 5 μl and 2 μl. Four dilution series were performed for each transfer volume and each assay, based on 6 concentrations and analyzed in qPCR tetraplicates, resulting in 16 qPCR datapoints per concentration per assay per volume.