I am performing RNA sequencing data analysis on weight loss data with three timepoints: before diet, 2 months after diet and 10 months after diet (I’m calling the timepoints TP1, TP2, TP3). My variable of interest is percentage weight lost, so I set my percentage weight lost at TP1 at 0 and TP2 and TP3 at (weight1-weight2)/weight1 and (weight1-weight3)/weight1 respectively. At the moment I am analysing my TP2vsTP1 and TP3vsTP1 separately. Not all individuals have data from all three timepoints.
My limma model looks like this:
design <- model.matrix(~ 0 + sequencing_center + sex + study_center + age + perc_weightlost_TP1TP2) v <- voom(dgeT1T2, design) corfit <- duplicateCorrelation(v,design,block=SubjectID) v <- voom(dgeT1T2,design,block= SubjectID,correlation=corfit$consensus) fit <- lmFit(v,design,block=ID,correlation=corfit$consensus) fit <- eBayes(fit) TP2vsTP1=topTreat(fit, adjust="BH", coef="perc_weight_TP1TP2", number=5000)
My interest is in finding out how gene expression changes according to percentage weight lost. Should I be putting the time point into my design matrix above? I had figured it was not necessary because the percentage weight lost already captures this information. However, if I want to put all three time points in my analysis, should I go about it as below?
design <- model.matrix(~ 0 + sequencing_center + sex + study_center + age + TP:perc_weightlost)
I would be grateful for any pointers on how to analyse this type of data.
> sessionInfo() R version 3.5.1 (2018-07-02) Platform: x86_64-pc-linux-gnu (64-bit) Running under: CentOS Linux 7 (Core) Matrix products: default BLAS: /apps/statistics2/R-3.5.1/lib/libRblas.so LAPACK: /apps/statistics2/R-3.5.1/lib/libRlapack.so locale:  LC_CTYPE=en_US.utf8 LC_NUMERIC=C  LC_TIME=en_US.utf8 LC_COLLATE=en_US.utf8  LC_MONETARY=en_US.utf8 LC_MESSAGES=en_US.utf8  LC_PAPER=en_US.utf8 LC_NAME=C  LC_ADDRESS=C LC_TELEPHONE=C  LC_MEASUREMENT=en_US.utf8 LC_IDENTIFICATION=C attached base packages:  stats graphics grDevices utils datasets methods base other attached packages:  edgeR_3.24.3 limma_3.38.3 loaded via a namespace (and not attached):  compiler_3.5.1 Rcpp_1.0.1 grid_3.5.1 locfit_1.5-9.1  statmod_1.4.30 lattice_0.20-38
University of Helsinki