Biological replicates and Diablo

[Lena]

Hi MaxBladen,
Thank you for your answer. I will try to clarify my design:

So, you have 4 to 5 repeated measurements for 16 different strains - so you have between 64 - 80 total samples?

Yes exactly, I have 78 samples

Is each sample independent or from the same individual at different time points?

Each of the five sample of a strain originally came from one unique strain. but they were each transplanted into a different bottle for growth before the experiment. They are of the same genotype but they are independent samples (each was measured only once during experiment)

Each of these strains (and therefore each sample) falls into one of two overarching classes and you’re concerned with the integrated differences between these two classes?

yes, for example all replicates of strains 1 to 9 (~45 samples) are in the first category and all replicates of strains 10 to 16 are in the other category. I would like to identify genes and metabolites that distinguish the two categories, but I would like to take into account the “non-independence” of the replicates in each category so as not to overestimate links between genes and metabolites for example.

The answer depends on what you want to do with it. However, I can’t think why you’d do this. withinVariation() works to decompose the input data - controlling for your repeated measurements of each strain. By doing them separately, the method won’t be able to control for this properly. I may be missing something here though.

Sorry for that, I tried to adapt this answer to my case withinVariation() on part of dataset - #2 by MaxBladen, but I understand now that it is not adapted, thank you
so taking into account your answers, I arrived at something like this:

rna<-t(read.table("rna",header=T,row.names=1)) #dim=78 600
metabolites<-t(read.table("metabolites",header=T,row.names=1)) #dim=78 400
samples_in_strains<-c(rep(1:4,each=5),rep(5,each=4),rep(6:15,each=5),rep(16,each=4))
design=data.frame(sp=samples_in_strains,row.names=rownames(rna))
categories<-c(rep("A",each=44),rep("B",each=34))
Y=as.factor(categories)
Xw_rna<-withinVariation(rna,design=design)
Xw_metabo<-withinVariation(metabolites,design=design)
X<-list(mRNA=Xw_rna,metabolite=Xw_metabo)
list.keepX<-list(mRNA=c(60,60),metabolite=c(30,30))
test<-block.splsda(X,Y,keepX=list.keepX)

However, on the PlotIndiv, my samples are not separated into two categories at all, everything is mixed up, do you think this is from the code or from my data?

Thank you !