In the paper published in the Open Access journal PLoS One, Jaume Roquer and the rest of the authors – all part of the Spanish Stroke Genetics Consortium – used three different commercial kits for DNA extraction for each sample, and then quantified the global DNA methylation (GDM) by a luminometric methylation assay (LUMA). In the 580 samples analysed, they found significant differences in GDM in the same samples between the three DNA isolation methods.
Large epidemiological studies, such as those carried out by the Spanish Stroke Genetics Consortium, are susceptible to accumulate variability by differences in the protocols, sample cohorts, reagent lots, and technologies used. This study demonstrates for the first time that the method of DNA extraction is indeed an important source of variability in LUMA methylation measurements.
The problem of this ‘batch effect’ becomes even more pronounced in collaborative studies, where different cohorts and differences in sample processing may threaten comparability of data and results. That is why the authors – which include Roberto Elosua’s group, also at the IMIM – recommend that methylation studies that apply multiple DNA extraction methods or in cross study comparisons should report the method used, and adjust their methylation results by this variable in order to avoid possible bias, be comparable and reach biologically meaningful conclusions.
Soriano-Tárraga C, Jiménez-Conde J, Giralt-Steinhauer E, Ois A, Rodríguez-Campello A, Cuadrado-Godia E, Fernández-Cadenas I, Montaner J, Lucas G, Elosua R, Roquer J, GeneStroke “The Spanish Stroke Genetics Consortium”. DNA Isolation Method Is a Source of Global DNA Methylation Variability Measured with LUMA. Experimental Analysis and a Systematic Review. PLoS One. 2013;8(4):e60750
Shigeru Kondo (Institute of Frontier Biosciences, Osaka University, Japan) gave one of the last talks at the “Computational approaches to networks, cells and tissues” meeting that took place this week at the PRBB Auditorium.
Co-organised by James Sharpe (CRG) and Hernán López-Schier (HZM), the meeting was supported by QuanTissue, a collaborative European network to bridge the gap between the traditional developmental cell biology, biophysics and systems biology. And so it did!
Most of the nearly 200 participants were physicysts or mathematicians, as one could tell from their presentations and posters full of complicated mathematical formulae. But the subjects they studied were all related to the development of tissues and organs within organisms.
Kondo, for example, talked about the pigmentation pattern of zebrafish and how the Turing model could explain it.
Although his lab found there is no actual diffusion of any molecules, they showed that the interaction between the two types of pigment cells that define the skin patterns in the fish can still be explained by the Turing reaction-diffusion model. Melanophores, one of the cell types, elongate long projections towards xanthophores, the other cell type, and the effect of this is mathematically equivalent to the classical Turing model. Interestingly, he showed how, changing one single gene his lab was able to generate fish with skin patterns resembling most of those present in nature, from leopards and jaguars to zebras. Hence, the title of this posts, with which he finished his talk: “If you want horses with spots or giraffes with stripes, I can make it!”.
The meeting is still going on – another two hours of good science if you rush!
A report by Maruxa Martinez, Scientific Editor at the PRBB