PT  - JOURNAL ARTICLE
AU  - Vicente A. Yépez M.
AU  - Laura S. Kremer
AU  - Arcangela Iuso
AU  - Mirjana Gušić
AU  - Robert Kopajtich
AU  - Eliška Koňaříkovà
AU  - Agnieszka Nadel
AU  - Leonhard Wachutka
AU  - Holger Prokisch
AU  - Julien Gagneur
TI  - OCR-Stats: Robust estimation and statistical testing of mitochondrial respiration activities using Seahorse XF Analyzer
AID  - 10.1101/231522
DP  - 2017 Jan 01
TA  - bioRxiv
PG  - 231522
4099  - http://biorxiv.org/content/early/2017/12/13/231522.short
4100  - http://biorxiv.org/content/early/2017/12/13/231522.full
AB  - Accurate quantification of cellular and mitochondrial bioenergetic activity is of great interest in many medical and biological areas. Mitochondrial stress experiments performed with Seahorse Bioscience XF Analyzers allow estimating 6 bioenergetics measures by monitoring oxygen consumption rates (OCR) of living cells in multi-well plates. However, detailed statistical analyses of OCR measurements from XF Analyzers have been lacking so far. Here, we performed 126 mitochondrial stress experiments involving 203 fibroblast cell lines to understand how OCR behaves across different biosamples, wells, and plates; which allowed us to statistically model OCR behavior over time. We show that the noise of OCR is multiplicative and that outlier data points can concern individual measurements or all measurements of a well. Based on these insights, we developed a novel statistical method, OCR-Stats, that: i) models multiplicative noise, ii) automatically identifies outlier data points and outlier wells, and iii) takes into account replicates both within and between plates. This led to a significant reduction of the coefficient of variation across experiments of basal respiration by 36% (P = 0.004), and of maximal respiration by 32% (P = 0.023). Also, we propose an optimal experimental design with a minimum number of well replicates needed to obtain confident results. Finally, we use statistical testing taking into account the inter-plate variation to compare the bioenergetics measures of two samples.