ScatterPlot with CummeRbund

Question

[Aynur]

Hello, I am new to bioinformatics and I am trying to get a scatterplot for my FPKM values, which I got from STAR-Cuffdiff pipeline. I also know that I need to use a different pipeline, but at the moment I have to work with this and get a scatterplot. So I used CummeRbund . The values are FPKM values from cuffdiff, control, and LIF group have 2 replicates. Cuffdiff gave me an average of the replicates. Here is the code I used.

library(cummeRbund), cuff<-readCufflinks(), s<-csScatterMatrix(genes(cuff)) I got the following plot. ![enter image description here][]

So, my question is 1. I am not able to add any arguments to this package. Where do I find other arguments available? 2. I want to color sig genes, how do I do that? 3. I want my scatter plot to be with linear fitting, is there any other program and package I get a better plot? Thanks a lot.

[antonioggsousa]

Hi,

Regarding the information, did you try to get information in the R console by typping: ?cuff or ?csScatterMatrix

For the csScatterMatrix you should get this information: https://www.rdocumentation.org/packages/cummeRbund/versions/2.14.0/topics/csScatter

Check also the vignette

António

[Aynur]

Hello António, Thank you very much for your answer. I followed the manual instructions.

a<-readCufflinks(system.file("extdata", package="cummeRbund")) #Create CuffSet object from sample data genes<-a@genes #Create CuffData object for all genes s<-csScatter(genes,'hESC','Fibroblasts',smooth=TRUE) #Create plot object s #render plot object My codes are below ,

library(cummeRbund) cuff<-readCufflinks('OSM_Cuffdiff') genes<- cuff@genes s<- csScatter(genes,'Control','OSM', colorByStatus = TRUE,smooth=TRUE) s And my data is below ![enter image description here][]

I want to color sig up and down genes please help me .

[antonioggsousa]

Can you print do this with your code and post here the result:

head(s$data)

Only then I can help more.

[Aynur]

Hello, Thanks.

`head(s$data).       Control       OSM.                                                      Gnai3|ENSMUSG00000000001 34.855600 37.116800                                             Pbsn|ENSMUSG00000000003   1.000000  1.000000. 
Cdc45|ENSMUSG00000000028 14.776100  8.918150.                                            H19|ENSMUSG00000000031    1.000000  1.000000.                                               `

I am supposed to have only gene names as row names, but...

[antonioggsousa]

Which genes or criteria you want to use to color your genes/points?

António

[Aynur]

I thought I will be able to access to P values and FDR , so I can color the ones with |log2(FC)| >=1 as significant up-regulated genes and down-regulated genes. But if I can merge another data frame, then I can have a separate significant gene list. Actually, I have a significant gene list that I want to color. Then the Q will be how to merge two data together and make a colored scatter plot.

Regards, Thank you for taking the time to help me out.

[antonioggsousa]

Yes, I agree with you. That is another question that you should make. Because as you can see the s$data holds only 2 columns with values for the Control and OSM variables. There isn't any FDR neither logFC there, so it's not possible to help you regarding this data.

I guess that the numbers being plotted are logCPM or similar, i.e., raw counts normalized. Do you know?

If you want to do the merging and plotting again, and this package does not provide that I would agree with you, make another question, perhaps referencing this post.

António

[Aynur]

I see. As this Cuffdiff-CummeRbund package, those values should be FPKM values (mean of two biological replicates). This cummeRbund is really not helpful.