Skip to main content

Table 3 The impact of mathematical modifications on model performance

From: Evaluation of absolute quantitation by nonlinear regression in probe-based real-time PCR

Model change

Mean R2 change

Intra-assay variation

Bias

Backgr. corrected data

-0.0065

+27%

+15%

Log10 transformation

0

+1%

-4%

Correction of baseline drift (linear)

+0.0001

+20%

-7%

Correction of baseline drift (Log10)

+0.0004

+35%

+12%

Weight early cycles (RAW)

-0.0006

+3%

+40%

Weight early cycles (Backgr. corrected)

+0.0056

+11%

+64%

Weight late cycles

-0.0006

+7%

+18%

  1. Performance of model modifications applied to different regression models, evaluated as single modifications. Even though four modifications show better or unchanged mean R2, all modifications actually add to the intra-assay variability.