Inquiry: p-values & q-values for cummeRbund volcano?

hyates · 04-27-2014, 10:13 PM

Hi,

Suppose I have the following command:

Code:

sigGene <- csVolcano(genes(cuffData), "A, "B")

It shows no differential expressed genes in the plot. It now can show some differential gene expression. Why? What changed? Is what being reported not taking into account the q-values which are adjusted?

Code:

sigGene <- csVolcano(genes(cuffData), "A, "B", alpha = 0.01, showSignificant =T)

Thanks,
GeekyOmega

PS - I looked at the bioconductor manual for cummeRbund. What is the default alpha for csVolcano()? Is it alpha = 0.05 for p-value/q-value? I can't figure out how it works exactly in the default/manual situation. Thanks.

sazz · 07-08-2014, 06:18 AM

Hello Hyates,

Depending on the output, I can say alpha is for q-value. But in the volcano plot it is the corresponding p-value. If your cutoff for q-value is 0.01 you should see that red dots for p-value starts around -log(2.42) in volcano plot (it would be exactly at -2.00 if alpha was accounted for p-value).

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
volcano plots cummeRbund	godzilla07	Bioinformatics	8	02-04-2015 01:55 PM
Are Cuffdiff q-values and p-values reliable when multiple treatments are analyzed?	gwilymh	Bioinformatics	1	03-16-2014 12:32 PM
cummeRbund diffData() or fpkm() values	jeales	Bioinformatics	0	06-26-2013 07:29 AM
CummeRbund csDendro error: "need finite ylim values" when replicates=TRUE	mebbert	Bioinformatics	4	07-18-2012 09:22 AM
beginning of quality values record not found & qual length () differs from seq length	papori	Bioinformatics	1	07-09-2012 01:25 PM

04-27-2014, 10:13 PM	#1
hyates Member Location: Kansas City Join Date: Jan 2014 Posts: 18	Inquiry: p-values & q-values for cummeRbund volcano? Hi, Suppose I have the following command: Code: sigGene <- csVolcano(genes(cuffData), "A, "B") It shows no differential expressed genes in the plot. It now can show some differential gene expression. Why? What changed? Is what being reported not taking into account the q-values which are adjusted? Code: sigGene <- csVolcano(genes(cuffData), "A, "B", alpha = 0.01, showSignificant =T) Thanks, GeekyOmega PS - I looked at the bioconductor manual for cummeRbund. What is the default alpha for csVolcano()? Is it alpha = 0.05 for p-value/q-value? I can't figure out how it works exactly in the default/manual situation. Thanks.

07-08-2014, 06:18 AM	#2
sazz Member Location: Istanbul, Turkey Join Date: Oct 2012 Posts: 28	Hello Hyates, Depending on the output, I can say alpha is for q-value. But in the volcano plot it is the corresponding p-value. If your cutoff for q-value is 0.01 you should see that red dots for p-value starts around -log(2.42) in volcano plot (it would be exactly at -2.00 if alpha was accounted for p-value).