Volcano Plot showing weird scatter

[Parashar]

Hi,

I have recently started with some RNA-seq analysis.
I used tuxedo suite for the analysis.

When I make a volcano plot using cummeRbund, I get weird alignment of my data-points parallel to X axis. Also the distribution of data-points is quite discrete and doesn't look to be continuous at higher -log p values.

I have attached a picture for reference.
http://seqanswers.com/forums/attachm...1&d=1394185690

Thanks

[TiborNagy]

Any other quality checking plots show some irregularities, or just this volcano plot?

[Parashar]

Hi TiborNagy

None of other plots like. CV2 plots, FPKM boxplots, scatter plots seem inconsistent. The anomaly between the QC steps conducted in the analysis of a different sample set and the current sample set lies only in Volcano plots.

[TiborNagy]

I guess somewhere in your pipeline there is a round/trunc step.

[Zapages]

Something like this happened with me. I had to go back to the cufflinks stage.

Please check all your cufflinks expression values for each individual steps. What this indicates is that one or more of samples is placed in the wrong group when you've have done cuffdiff. As some gtf files have mistakenly switched up.

Please try to do cuffmerge again with the correct order of the gtf files. Then run Cuffdiff again.

I hope this helps.