Hello,
my fellow biologist is using nsolver for analyzing nanostring data. One of the features of nsolver that she loves is it provides with signature scores for pathways and boxplots for these scores comparing sample types, scatter plots, etc.
The problem is: there is no confidence scores in these plots or these results, and she asked me to compute those confidence scores for her. How could I do so? I think nsolver uses pathifier or something similar to it, but I am not sure. Pathifier does not provide with p values either. What do you recommend? I downloaded the results from nsolver and they look like this:
Cell Cycle - Apoptosis Chromatin Modification DNA Damage - Repair Driver Gene Hedgehog
Sample1.RCC 6.874940011 1.575197879 3.019260549 4.410085616 3.165996225
Sample2.RCC 0.797363391 -0.559763334 -0.63705548 1.31730206 -0.778248895
Sample3.RCC -2.252367228 -0.762525873 -0.931860336 -1.774930444 -0.667779337
Sample4.RCC -1.028090843 -0.05319562 -2.242034205 0.292504896 2.065701421
These are the signature scores. Then, I also have this information about in which pathways is located each gene :
Log2 fold change BONF.p.value Gene.sets
FGF10-mRNA -1.8 0.000191 MAPK, PI3K, Ras
CDC6-mRNA 1.25 0.00466 Cell Cycle - Apoptosis
PIK3CA-mRNA 0.461 0.0155 Cell Cycle - Apoptosis, Driver Gene, JAK-STAT, PI3K, Ras
Any ideas on how to get a confidence value for the signature scores? nanostring support said I should do a t-test, but I dont know how! using what data?
Thank you.
