Hi,

I am trying to address the question which genes associate with change of the metabolite levels following the diet. I have paired samples (before and after the diet). I have previously asked about possible adjustments to the paired set (https://support.bioconductor.org/p/134040/#134068) so just few follow up questions

That is an example of my data organization.

Sample  ID  Timepoint   Metabolite  Log_Metabolite  Sex Age BMI
1a  1   1    10.3   3.36    1   56.5    38.1
1b  1   2    20.1   4.33    1   57.7    27.2
2a  2   1    11.0   3.46    2   21.0    44.0
2b  2   2    28.7   4.84    2   22.2    25.8
3a  3   1    12.4   3.63    1   30.0    33.1
3b  3   2    65.8   6.04    1   31.3    30.0
4a  4   1   112.0   6.81    1   67.0    31.5
4b  4   2   100.7   6.65    1   68.0    29.7
5a  5   1    36.2   5.18    1   53.5    36.8
5b  5   2    89.1   6.48    1   54.5    32.9
6a  6   1    12.9   3.69    2   25.7    40.4
6b  6   2    29.0   4.86    2   26.7    37.6
7a  7   1    15.1   3.92    2   44.8    35.7
7b  7   2    98.2   6.62    2   45.9    23.1
8a  8   1     8.0   3.00    1   25.4    29.9
8b  8   2    11.6   3.53    1   26.6    24.8

1) I would like to ask if it is OK to adjust paired data for the RNA quality (RIN) value. My supervisor insists on including some technical factors (at least a RIN)

design <- model.matrix( ~ ID + RIN + Timepoint)

2) Gordon commented previously

"Even if you included Metabolite in a completely different non-paired analysis, the Metabolite concentration would need to be a log-scale. Taking differences of unlogged Metabolite concentrates (as you have to get the change variable) is not a meaningful thing to do."

I understand that this is not edgeR but rather statistics related question, but I'm trying to understand why that's the case. The statistician in my university told me that in lm I should use my dependent variables (like glucose and insulin levels, BMI or said Metabolite) in the "raw" from- meaning untransformed.

Thank you very much for your help!