RNA-Seq differential gene expression without biological and technical replicates.

Question

[utsafar]

What I have is six RNA-Seq datasets from three developmental stages in two conditions (control and treatment), without biological and technical replicates. (I did all the following in usegalaxy.eu) Each dataset was mapped to reference genes using tophat. htseq-count was used to count aligned reads. I tried to use DESeq2 for differential expression analysis but it works only when I have replicates. Is there any way/tool to do DE without biological and technical replicates? suggestions considering Galaxy friendly tools are preferred. Thank you so much

Note: The data belongs to a friend who said that they have no replicates because of budget shortages.

[GenoMax]

If you wish to get galaxy specific advice you may want to post this over in Galaxy help forum: https://help.galaxyproject.org/

Answer 1

[sambhav2mishra]

You can try using edgeR package (go through edgeR manual) , it has an explanation where you can analyse DGE without any replicates. Thanks

Answer 2

[Ido Tamir]

How to run the Deseq2 tool without replicates. maybe look at NOI-Seq http://seqanswers.com/forums/showpost.php?p=107433&postcount=2

or just do it in excel: sort by ratio from high to low, same reliability (but no normalisation). The result is not to be trusted. It would have been better to use just 1 developmental stage or only control or treatment in a time course. Then the data would have been at least somehow reliable. RNA-Seq data is noisy and you simply don't know which of the > 20000 data points represents a reliablel difference without replicates.

Answer 3

[eyonesi]

Deseq, edgR and bayseq are pakages that can analyze without replicate. But, i think, they require replicate when running in the galaxy. Regardless of the galexy, edgR and Bayseq can be better choices for anlyses without replicat.

Answer 4

[a410638758]

You can use NOIseq or edgeR to process samples without biological and technical replicates