Comparisons of parametric and non-parametric methods for analyzing RT-PCR experiment data

The real-time reverse-transcript polymerase chain reaction (RT-PCR) test is a widely used laboratory technique that is highly sensitive and reliable for measuring the quantification of gene expression levels and diagnosing various of diseases, including COVID-19. The RT-PCR experiments often have correlated technical replicates of a small number of samples. However, current statistical analysis of RT-PCR assumes a large sample size and does not account for correlated structure across the replicates. In this paper, we review popular statistical methods for analyzing RT-PCR data and propose a permutation method that accounts for the small sample size and the correlated structure of RT-PCR data. Our proposed method provides a more accurate and efficient analysis of RT-PCR data. We provide an R program to implement our method for practitioners. •Current statistical analysis of RT-PCR assumes a large sample size and does not account for correlated structure across the replicates.•We propose a permutation method that accounts for both the small sample size and replicated structure of the RT-PCR data.•We have created a Github website where practitioners can obtain an R program used in this paper.