I have a long list of protein expression with different factors.
I did multiple ANOVA on it to find wich protein were differentialy expressed.
#Doing one anova by protein.
donnee%>% group_by(Protein) %>% do(tidy(aov(Norm ~ Cultivar*Treatment/Strain, data = .))) ->AOV.Norm_Treatment_cultivar
It gave me :
Protein term df sumsq meansq statistic p.value
1 TraesCS1A01G002200.1 Cultivar 2 0.0628 0.0314 0.890 0.424
2 TraesCS1A01G002200.1 Treatment 1 0.418 0.418 11.8 0.00214
3 TraesCS1A01G002200.1 Cultivar:Treatment 2 0.0179 0.00896 0.254 0.778
4 TraesCS1A01G002200.1 Cultivar:Treatment:Strain 6 0.348 0.0580 1.64 0.179
5 TraesCS1A01G002200.1 Residuals 24 0.847 0.0353 NA NA
6 TraesCS1A01G005600.3 Cultivar 2 0.304 0.152 2.55 0.0994
Now I would like to find wich one a differentially expressed when exposed to a precised strain.
To do that I did a new ANOVA on my prot and a wnated to do multiple TUKEY on it, but I just can't.
significative_data_%>% group_by(Protein) %>% do(tidy(aov(Norm ~ Strain, data = .))) ->aov_significative
aov_significative%>% do(tidy(HSD.test(aov_significative,"Strain",group=TRUE,console=TRUE))) ->list_of_prot
It print :
rror in HSD.test(aov_significative, "Strain", group = TRUE, console = TRUE) :
"MSerror" argument is missing
In mean.default(junto[, 1]) :
argument is not numeric or logical: returning NA
So tukey doesn't work or I use it wrong.
Any idea to help me (by a new path or just correct this one)