Hello,
I was looking at a recent paper that use S. pombe spike-in to normalize their S. cerevisiae RNA-Seq data (Rodrıguez-Molina et al., 2016, Molecular Cell 63, 433–444). They said that they used a LOESS regression to normalize their data between samples in order to make the claim of nascent RNA transcription having a 4-fold decrease between WT and mutant situations. Rodrigez-Molina et al. paper cited (Loven, J. et al. (2012). Revisiting global gene expression analysis. Cell 151, 476-482).
I don't understand fully how they used LOESS to determine the amount to normalize according to this paper. According to the Loven paper, S. pombe RPKM values from each gene from two samples are plotted on X and Y axis and calculate the LOESS curve. So, they have the function that fits between the two S. pombe reads, is the best smoothing value the one they use to normalize their S. cerevisiae reads?
From Rodriguez-Molina et al. (2016), "Annotation files were obtained from the same databases from where the genome sequences used in initial read alignment were downloaded. S. cerevisiae read counts were normalized with loess regression using S. pombe read counts to fit the loess as described by (Loven et al., 2012)."
Thanks
I was looking at a recent paper that use S. pombe spike-in to normalize their S. cerevisiae RNA-Seq data (Rodrıguez-Molina et al., 2016, Molecular Cell 63, 433–444). They said that they used a LOESS regression to normalize their data between samples in order to make the claim of nascent RNA transcription having a 4-fold decrease between WT and mutant situations. Rodrigez-Molina et al. paper cited (Loven, J. et al. (2012). Revisiting global gene expression analysis. Cell 151, 476-482).
I don't understand fully how they used LOESS to determine the amount to normalize according to this paper. According to the Loven paper, S. pombe RPKM values from each gene from two samples are plotted on X and Y axis and calculate the LOESS curve. So, they have the function that fits between the two S. pombe reads, is the best smoothing value the one they use to normalize their S. cerevisiae reads?
From Rodriguez-Molina et al. (2016), "Annotation files were obtained from the same databases from where the genome sequences used in initial read alignment were downloaded. S. cerevisiae read counts were normalized with loess regression using S. pombe read counts to fit the loess as described by (Loven et al., 2012)."
Thanks