This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Rna-Seq Within Group Variability

We have RNA-seq data for 6 individuals showing one phenotype and 6 showing another (12 total samples). Before doing a diff. analyses between two groups of 6 each, we'd like to assess the variability within each group using the FPKM values given by cufflinks. Can someone point out a good method? Do EdgeR / DESeq etc. do this? All suggestions welcome.

I think in the microarray world there are several tests available for this, and I am just wondering if there is anything analogous for RNA-seq counts data.

Thanks, forum.

rna
• 4,960 views • link
updated by Sean Davis Sean Davis • written by learnerforever learnerforever

what are the standard within-sample metrics for microarrays?

• 0 views • link
written by Jeremy Leipzig Jeremy Leipzig

2 answers

You might try simple heatmaps, hierarchical clustering, and PCA (or MDS) plots of all 12 samples. These types of diagnostic plots often give a good sense of sample relationships and the within- and between-group variances. DESeq can also be useful for this process.

Note that DESeq and edgeR are going to take raw counts and not FPKM values.

• 0 views • link
written by Sean Davis Sean Davis

If they take raw counts; how do they handle the variability in initial reads for various samples?

• 0 views • link
written by Rm Rm

I use simple diagnostic plots to do that, like MA plot or simple dot plot. I think DESeq/edgeR both have specific functions for doing the MA plots and they work quite well. DESeq also has functions for extensive variance analysis (a bit involved in terms of statistics), for which you may see the manual of DESeq.

• 1 views • link
written by Vitis Vitis

Log in to answer this question.

More posts like this