Djemail Ismaili1, Bastiaan Geelhoed2, Torsten Christ3. 1. Institute of Experimental Pharmacology and Toxicology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; Department of Cardiology, University Heart and Vascular Center, Hamburg, Germany; DZHK (German Centre for Cardiovascular Research), Partner Site Hamburg/Kiel/Lübeck, Hamburg, Germany. 2. Department of Cardiology, University Heart and Vascular Center, Hamburg, Germany; DZHK (German Centre for Cardiovascular Research), Partner Site Hamburg/Kiel/Lübeck, Hamburg, Germany. 3. Institute of Experimental Pharmacology and Toxicology, University Medical Center Hamburg-Eppendorf, Hamburg, Germany; DZHK (German Centre for Cardiovascular Research), Partner Site Hamburg/Kiel/Lübeck, Hamburg, Germany. Electronic address: t.christ@uke.de.
Abstract
OBJECTIVES: Variability of ion currents is major issue when used for significance testing. One of the simplest approach to reduce variability is normalization to cell membrane size. However, efficacy of Ca2+ currents (ICa) normalization is unknown. Beside absolute variability, the type of distribution since non-Gaussian distribution makes application of nonparametric test necessary. METHODS: We retrospectively analyzed individual ICa amplitudes measured in ventricular cardiomyocytes from mice, rats and humans and in atrial cardiomyocytes from humans in sinus rhythm and in atrial fibrillation. ICa was normalized to cell membrane size, estimated from capacitance transients. In addition, data were Log transformed to reach Gaussian distribution. Normalized and transformed data were analyzed for variability and applicability of parametric vs. nonparametric tests. RESULTS: There was strong correlation between ICa and cell membrane size. However, correlation coefficient was rather low. Normalizing ICa had an inconsistent effect on variability. Variability of ICa in cells from the same patient/animal was not different cardiomyocytes from humans, rat and mice. Calculation of mean values based on mean values of cells from individuals (patients or animals) vs. mean values calculated for all cells drastically reduces statistical power to detect differences between the groups. Log transformation of ICa allowed application of much higher sensitive parametric testing, compensating loss of power. CONCLUSION: Impact of cell membrane size to ICa is low and may limit efficacy of normalization of ICa to reduce variability. In contrast, Log transformation of ICa data reduces variability and can increase statistical power to detect difference between ICa datasets.
OBJECTIVES: Variability of ion currents is major issue when used for significance testing. One of the simplest approach to reduce variability is normalization to cell membrane size. However, efficacy of Ca2+ currents (ICa) normalization is unknown. Beside absolute variability, the type of distribution since non-Gaussian distribution makes application of nonparametric test necessary. METHODS: We retrospectively analyzed individual ICa amplitudes measured in ventricular cardiomyocytes from mice, rats and humans and in atrial cardiomyocytes from humans in sinus rhythm and in atrial fibrillation. ICa was normalized to cell membrane size, estimated from capacitance transients. In addition, data were Log transformed to reach Gaussian distribution. Normalized and transformed data were analyzed for variability and applicability of parametric vs. nonparametric tests. RESULTS: There was strong correlation between ICa and cell membrane size. However, correlation coefficient was rather low. Normalizing ICa had an inconsistent effect on variability. Variability of ICa in cells from the same patient/animal was not different cardiomyocytes from humans, rat and mice. Calculation of mean values based on mean values of cells from individuals (patients or animals) vs. mean values calculated for all cells drastically reduces statistical power to detect differences between the groups. Log transformation of ICa allowed application of much higher sensitive parametric testing, compensating loss of power. CONCLUSION: Impact of cell membrane size to ICa is low and may limit efficacy of normalization of ICa to reduce variability. In contrast, Log transformation of ICa data reduces variability and can increase statistical power to detect difference between ICa datasets.
Authors: Djemail Ismaili; Katrin Gurr; András Horváth; Lei Yuan; Marc D Lemoine; Carl Schulz; Jascha Sani; Johannes Petersen; Hermann Reichenspurner; Paulus Kirchhof; Thomas Jespersen; Thomas Eschenhagen; Arne Hansen; Jussi T Koivumäki; Torsten Christ Journal: Cells Date: 2022-08-05 Impact factor: 7.666
Authors: Frederik Flenner; Christiane Jungen; Nadine Küpker; Antonia Ibel; Martin Kruse; Jussi T Koivumäki; Anna Rinas; Antonia T L Zech; Alexandra Rhoden; Paul J M Wijnker; Marc D Lemoine; Anna Steenpass; Evaldas Girdauskas; Thomas Eschenhagen; Christian Meyer; Jolanda van der Velden; Monica Patten-Hamel; Torsten Christ; Lucie Carrier Journal: J Mol Cell Cardiol Date: 2021-05-03 Impact factor: 5.000