Validation of an algorithm for automatic quantification of nucleic acid copy numbers by real-time polymerase chain reaction

Real-time quantitative polymerase chain reaction (PCR) with on-line fluorescence detection has become an important technique not only for determination of the absolute or relative copy number of nucleic acids but also for mutation detection, which is usually done by measuring melting curves. Optimum assay conditions have been established for a variety of targets and experimental setups, but only limited attention has been directed to data evaluation and validation of the results. In this work, algorithms for the processing of real-time PCR data are evaluated for several target sequences (p53, IGF-1, PAI-1, Factor VIIc) and compared to the results obtained by standard procedures. The algorithms are implemented in software called SoFAR, which allows fully automatic analysis of real-time PCR data obtained with a Roche LightCycler instrument. The software yields results with considerably increased precision and accuracy of quantifications. This is achieved mainly by the correction of amplification-independent signal trends and a robust fit of the exponential phase of the signal curves. The melting curve data are corrected for signal changes not due to the melting process and are smoothed by fitting cubic splines. Therefore, sensitivity, resolution, and accuracy of melting curve analyses are improved.