cat("A paragraph about Limma's method.\nLa variabilité systématique des données est modélisée avec une approche linéaire pour la différencier de la variabilité aléatoire. Cette modélisation linéaire est très similaire à l’ANOVA classique ou à la régression multiple, sauf qu’un modèle est adapté à chaque pic.\n")
cat("For the differential marking/binding analysis we use the limma approach to RNA-seq [@ritchie2015]. Read counts are converted to log2-counts-per-million (logCPM) and the mean-variance relationship is modelled either with precision weights or with an empirical Bayes prior trend. Here we use the the precision weights approach called “voom” [@law2014]. This transformation permits to apply the linear modelling in the limma package can be applied to sequencing data. The systematic variability of the data is modeled with a linear approach to differentiate it from the random variability. This linear modeling is very similar to classical ANOVA or multiple regression, except that a model is adapted to each peak.")