Question: Barcodeplot limma function using two logFC lists
0
gravatar for srmaruyama
21 months ago by
srmaruyama0
srmaruyama0 wrote:

Dear All,

To visualize my gene set enrichment analysis, I have been using $table$logFC from an object of class "DGELRT" as statistics argument together with two indices (index, index2). Now, I am interested in to visualize one index using two $table$logFC lists, to display a graph with the same plot configuration as presented by Ng et al., (Plos Genet. 2015) in Figure 2B . How could be the code to do the plot in that way?

Thank you,

Sandra

ADD COMMENTlink modified 21 months ago by Gordon Smyth38k • written 21 months ago by srmaruyama0

I'm not quite sure what you are trying to do. Are you trying to assess the correlation of the two logFC lists? A barcode plot can certainly do that, and in fact that's what Figure 2B of Ng et al is doing. But, if that's what you want, what is the "one index" you are refering to then?

Are the two logFC lists different comparisons for the same data set, or are they from different data sets?

ADD REPLYlink modified 21 months ago • written 21 months ago by Gordon Smyth38k

Dear Gordon,

The two logFC lists are different comparisons from the same data set. I have 3 groups, let's say A, B and C. My two lists are AvsB$table$logFC and BvsC$table$logFC and I used a gene set list (index) from MSigDB C3 collection (TFT transcription factor targets). What I observed is that, index genes are regulated in an opposite way depending on the logFC lists, e.g., gene X is upregulated in AvsB, whereas in BvsC is downregulated and vice-versa.

I was wondering if a barcodeplot can show this correlation of the two comparisons in one plot. As the gene set list (index) is transcription factor targets, there is no UP or DOWN atributes to this gene set for doing 2 indices. If I break down the gene set in genes UP (index1), genes DOWN (index2) based on logFC values from AvsB for example, I will come up with a plot showing only AvsB comparison and the other way round for BvsC. What I thought when I saw the Figure 2B of Ng et al, was that Ts65DnErg+/+/+ and Erg+/+  labels at side of the plot were different comparisons (different LogFC lists), but now I think that I misunderstood that barcodeplot. 

I will follow your suggestion about to do indices from the second logFC list.

Thank you very much for your attention.

Sandra

ADD REPLYlink written 21 months ago by srmaruyama0
Answer: Barcodeplot limma function using two logFC lists
0
gravatar for Gordon Smyth
21 months ago by
Gordon Smyth38k
Walter and Eliza Hall Institute of Medical Research, Melbourne, Australia
Gordon Smyth38k wrote:

One way to proceed is the define the indices from the second logFC list. You can use:

barcodeplot(logFC, index1, index2)

where logFC is the first logFC list, index1 identifies those genes that are significantly up in the second list, and index2 identifies those genes that are significantly down the second list.

Does this answer your question?

ADD COMMENTlink written 21 months ago by Gordon Smyth38k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 166 users visited in the last hour