Analyzing barcode sequencing experiments with DESeq2/edgeR
Entering edit mode
nc2849 • 0
Last seen 19 months ago
United States

Is it possible to analyze data from barcode screens using DESeq2 (or a similar package, such as edgeR)?

More details: I have a library of barcoded knockout strains (much like an shRNA or CRISPR library) so that I can look for fitness differences as measured by strain count changes after exposure to the experimental condition. After growth in liquid culture, I took an aliquot of the strain as the input and treated several aliquots under my experimental conditions.

My understanding is that DESeq2 is more powerful when you have replicates of your control condition (e.g. 3 control vs 3 treated). However, in my case my control is the input "ground truth" of the prevalence of strains before treatment. I have several aliquots of the input, but they are derived from the same input sample. Would it be possible to consider these aliquots as different biological replicates for DESeq2 in order to power the analysis?

Another strategy I've seen is to use separate library inputs for each experimental replicate (so that you have e.g. 3 paired control-treated samples). Is there a way of telling DESeq2 to match these? Or would another analysis package be more appropriate to use in this case?

edgeR DESeq2 • 924 views
Entering edit mode
Last seen 2 hours ago
WEHI, Melbourne, Australia

Dai Z, Sheridan JM, Gearing, LJ, Moore, DL, Su, S, Wormald, S, Wilcox, S, O'Connor, L, Dickins, RA, Blewitt, ME, Ritchie, ME (2014). edgeR: a versatile tool for the analysis of shRNA-seq and CRISPR-Cas9 genetic screens. F1000Research 3, 95.

Entering edit mode
chaco001 • 0
Last seen 15 months ago
United States

I went through a similar tool decision process. I think the problem is like you described--there isn't an obvious way to have a common T0 sample and then a second "level" of differential analysis of the fitnesses.

I think the best solution if you want to use off-the-shelf tools is to calculate log2-fold changes (aka fitnesses) of your strains relative to your T0. Then, these can be used as input into limma to calculate differential fitness between your treatments.

Or, you could use DESeq2/edgeR to calculate differential fitnesses among your treatments while ignoring the T0, but as you know the problem here is that you don't know the "baseline" fitnesses.


Login before adding your answer.

Traffic: 460 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6