Check that replicates are good

[Blahah404]

Lots of people also do grid plots of Pearson/Spearman correlation, sometimes with a hierarchically clustered dendrogram on the axes. Also easy to do in R with the cor() function once you have a matrix of genes by samples.

[colindaven]

You can also just do a scatter plot, for example in R, of Rep1 vs Rep2. This can be a bit easier to understand when communicating your results to other researchers.

[dariober]

Hi,
I'm new to RNA-seq analysis. I have biological duplicates of 4 different conditions. I want to check how good the replicates are for each condition. I have got them mapped to the genome and have run cuffdiff on the files, but I was wondering what comparisons people do/tools people use to check their replicates?
Thanks

Hi- A plot of principal components (PCA plot) and/or multidimensional scaling (MDS plot) and/or a dendrogram are usually quite helpful in spotting sample outliers and to check that the replicates cluster together. Once you have a matrix of genes and samples they are quite easy to generate in e.g. R. See also the vignettes in edgeR and DESeq (R/bioconductor packages).

Dario

[jake13]

Hi,
I'm new to RNA-seq analysis. I have biological duplicates of 4 different conditions. I want to check how good the replicates are for each condition. I have got them mapped to the genome and have run cuffdiff on the files, but I was wondering what comparisons people do/tools people use to check their replicates?
Thanks