Optimal use of statistical methods to validate reference gene stability in longitudinal studies

Abstract

Multiple statistical approaches have been proposed to validate reference genes in qPCR assays. However, conflicting results from these statistical methods pose a major hurdle in the choice of the best reference genes. Recent studies have proposed the use of a minimum of three different methods but there is no consensus on how to interpret conflicting results. Researchers resort to averaging the ranks or attributing a weighted rank to candidate genes. However, we report here that the suitability of these validation methods can be influenced by the experimental setting. Therefore, averaging the ranks can lead to suboptimal assessment of stable reference genes if the method used is not suitable for analysis. As the respective approaches of these statistical methods are different, a clear understanding of the fundamental assumptions and the parameters that influence the reference gene stability calculation is necessary. In this study, the stability of 10 candidate reference genes (Actb, Gapdh, Tbp, Sdha, Pgk1, Ppia, Rpl13a, Hsp60, Mrpl10, Rps26) was assessed using four common statistical approaches (GeNorm, NormFinder, Coefficient of Variation or CV analysis and Pairwise ΔCt method) in a longitudinal experimental setting. We used the development of the cerebellum and the spinal cord of mice as a model to assess the suitability of these statistical methods for reference gene validation. GeNorm and the Pairwise ΔCt were found to be ill suited due to a fundamental assumption in their stability calculations. Highly correlated genes were given better stability ranks despite significant overall variation. NormFinder fares better but the presence of highly variable genes influences the ranking of all genes because of the algorithm's construct. CV analysis estimates overall variation, but it fails to consider variation across groups. We thus highlight the assumptions and potential pit-falls of each method using our longitudinal data. Based on our results, we have devised a workflow combining NormFinder, CV analysis along with visual representation of mRNA fold changes and one-way ANOVA for validating reference genes in longitudinal studies. This workflow proves to be more robust than any of these methods used individually.

RNA-Seq is not required to determine stable reference genes for qPCR normalization

“…For the sciatic nerve dataset, the conventional mouse reference genes chosen were Actb, Gapdh,Tbp,Sdha,Pgk1,Ppia,Rpl13a,Hsp60,Mrpl10,Rps26. These mouse reference genes have been previously used to establish the qPCR data analysis workflow (Sundaram et al, 2019) Target gene selection: Differentially expressed genes that exhibited Padj<0.05 were first retained from the results dataframe. Next, genes that exhibited log2FC values above +0.6 and below -0.6 were retained and partitioned into 2 separate lists.…”

“…In such cases, one outlier Cq was removed to have at least duplicate Cq values for each biological sample and an SD < 0.20. Reference gene validation was performed according to our qPCR data analysis workflow (Sundaram et al, 2019). Visual representation of potential intrinsic variation in reference genes was identified by plotting the raw expression profiles (2 -ΔCq ) of all candidate reference genes.…”

“…In the last two decades, many statistical approaches have been proposed to help researchers identify stable reference genes from a given set of candidates (Vandesompele et al, 2002;Pfaffl et al, 2004;Andersen et al, 2004;De Spiegelaere et al, 2015;Silver et al, 2006;Boda et al, 2009). However, the calculations and assumptions employed in these approaches fundamentally differ and could give rise to conflicting results (Sundaram et al, 2019). To address these concerns, we recently reviewed these approaches and following a comparison of existing methods including Coefficient of Variation (CV) analysis, GeNorm, NormFinder, pairwise ΔCT method and Best Keeper, we proposed an effective qPCR analysis workflow (Sundaram et al, 2019).…”

“…However, the calculations and assumptions employed in these approaches fundamentally differ and could give rise to conflicting results (Sundaram et al, 2019). To address these concerns, we recently reviewed these approaches and following a comparison of existing methods including Coefficient of Variation (CV) analysis, GeNorm, NormFinder, pairwise ΔCT method and Best Keeper, we proposed an effective qPCR analysis workflow (Sundaram et al, 2019). Our approach combining visual representation and statistical testing of intrinsic variation, CV analysis for identifying overall reference gene variation and the NormFinder algorithm, proved to be effective in determining stable reference genes from a given set of candidates.…”

Evaluation of suitable reference genes for normalization of quantitative real-time PCR analysis in rice plants under Xanthomonas oryzae pv. oryzae--infection and melatonin supplementation

“…Theoretically, reference genes should remain stable under different experimental conditions and may show the same mRNA level in all type of cells and tissues. However, there is no universal internal standard gene that fulfills completely this criterium (Sundaram et al 2019). Hence, the validation of the expression stabilities of reference genes is necessary for the accurate acquisition of qRT-PCR data.…”