I have RNASeq data, without replicates. Has anyone used the ISoDE method for analysis? Here's the reference of the paper
S. Al Seesi and Y.T. Tiagueu and A. Zelikovsky and I.I. Mandoiu, Bootstrap-based differential gene expression analysis for RNA-Seq data with and without replicates, BMC Genomics 15(Suppl 8):S2 , 2014
In theory, classic edgeR and DESeq can be used without replicates. They were designed when RNAseq was even more expensive than now, and not having replicates was quite common. Perhaps you will need to select some specific parameters for dispesion estimation in those tools. Still, you need to have a look at the final p.adj=FDR to check if you will get statistically sound results.
Differential Expression of RNASeq datasets can be carried out using NOISeq (if you have technical replicates or even not have any replicates) and NOISeqBio (for biological replicates). Both are the part of NOISeq Bioconductor package of R.
The analysis can be performed using either raw counts (no. of reads mapped to the particular feature) or normalized values (RPKM, FRKM etc.).
Many quality control and differential expression analysis plots such as GC Content, length bias, biotype (if you have biotype information along with your gene features) , Scatter plot, Expression plot (MD plot) etc. can be generated to visualize.
Hope this helps!
There are other tools also that can be used for RNA-Seq data without replicates such as EBSeq and GFOLD.
Now, GFOLD has been made exclusively for RNASeq data without replicates and has been recommended by a lot of bloggers.
Now, I also prefer EBSeq because I don't have to specify the variability in replicates will be which has to be done in NOISeq, edgeR and DESeq.
EBSeq predicts the variability itself and finds DEGs :)
But, I would recommend that you run your data from all these tools and report commonly found DEGs across all the tools.
Hi, by now I am using EBseq HMM that is a sort of extension of EBseq for time course experiments. I have an experiment with single replicates and two contrasting genotypes. This program clasifies genes into expression paths. I have identified that this program ranks the genes with a higher probability if they present great changes between each time point... this might work if you also have a time course experiment.
@Kaalindi Misra
@Diana Duarte-Delgado
- Hi, can you please help me with links to tutorials for using GFOLD and EBSeq?
Hi, this is the EBseq manual
http://bioconductor.org/packages/release/bioc/vignettes/EBSeq/inst/doc/EBSeq_Vignette.pdf
Good luck!
- Badges
- Science topic
- Similar topics
- Biological Science
- Genetics
- Gene Expression