RNA-seq Reproducibility analysis

[lwhitmore]

Hey,
I have 3 RNA-seq biological replicates of which I want to assess the reproducibility between the 3 replicates. I am only planning on performing HTSeq and looking at the read counts for our genes of interest. All the reproducibility analyses I have found are designed to analyze reproducibility based on differential expression between replicates. Are there any methods designed to look at the reproducibility based on the raw read counts between replicates?

Hopefully this question makes sense and any help anyone could provided would be great

[aggp11]

As a quick analysis you could look at scatter plots of the read counts (hopefully normalized) and also perform some clustering analysis on the same to see reproducibility of your libraries.

[linusvanpelt]

I have been working on this also lately and my solution so far is to look at agreement between the replicates using Bland-Altman plots (plotting mean vs. difference) . I use a 1-sample t-test to check if the mean value of the difference significantly differs from 0. Obvisously, you can also reduce this only to your genes of interest.

Cheers,
Tobias