SEQanswers

SEQanswers (http://seqanswers.com/forums/index.php)
-   Bioinformatics (http://seqanswers.com/forums/forumdisplay.php?f=18)
-   -   using DESeq2 for RNA-Seq analysis (http://seqanswers.com/forums/showthread.php?t=34718)

ogla 10-23-2013 12:03 AM

using DESeq2 for RNA-Seq analysis
 
Hello All,
i have used the DESeq package for RNA-Seq analysis up until recently.
after discovering the DESeq2 package, i have tried using it for that kind of analysis, and encountered some problems.
1. i have read that the package offers two different statistical testing for differential expression : the "nbinomWalsTest" and the "nbinomLRT", i can't seem to understand fully what is the difference between them and what is the recomanded test using for RNA-Seq analysis.
2. my current data contains 4 conditions, one of them is the control and the other three ate different treatments. when using the nbinomWaldTest i have used the "contrast" option in order to define the specific comparison that i am intreasted in. is there some kind of option like that for the LRT test? since right now i only receive results comparing all conditions to the control.
3. i would like to receive MA plots for my comparisons between different conditions, right now all i was able to receive is the plots of comparing each condition to the control (since those are the results available in the results() function by default)

i general, if anyone has expreience using this tool on RNA-Seq analysis, i would really appritiate any help regarding the steps to take for receiving full results.

thank you very much,
olga.

marianaboroni 02-18-2014 06:43 AM

Dear Olga,

Did you hear back from someone ?
I'm also interested in the answers of your questions...

ogla 02-19-2014 11:38 PM

No, actually i did not hear back from anyone.
I am using the Wald test for the analysis at this point.
If i will recieve any information regarding those issues i will post the answers here.

Wolfgang Huber 02-20-2014 06:13 AM

Olga, Mariana

sorry for not responding earlier, not sure why this slipped through (it might also be useful to ask questions about Bioconductor packages on the Bioconductor mailing list http://www.bioconductor.org/help/mailing-list/ first).

A lot has changed (improved) in the user interface and documentation of the DESeq2 package since October, so I suggest you have a look at the latest version (http://www.bioconductor.org/packages...ml/DESeq2.html ) and the vignette Analyzing RNA-Seq data with the "DESeq2" package. Also, we posted a preprint of the DESeq2 paper: http://biorxiv.org/content/early/2014/02/19/002832
Let us then know when questions remain!

Kind regards
Wolfgang

rajarapupriya 03-19-2014 05:40 AM

Dear Wolfgang,

I am trying to use DESeq2 as I have seen many changes from the previous version. I am having a lot of trouble following the data (raw read counts) input part.
In the vignette it mentions - " An example of the steps to produce a SummarizedExperiment can be found in the data package parathyroidSE,", but the parathyroidSE link seems non functional.
Could you please fix this error or direct me to a source where I can get help getting started with DESeq2.

Thanks
Priya

Michael Love 03-20-2014 04:23 AM

Just to note that the conversation with Olga continued on the Bioc mailing list:

http://permalink.gmane.org/gmane.sci...onductor/53790

Priya, the link is fixed in the development version, which will be released in < 1 month.

the link should have been:

http://bioconductor.org/packages/rel...thyroidSE.html

We have created another vignette for DESeq2 version 1.3 which goes slower, starting with how to count reads in genes. This should appear here in 1-2 days:

http://www.bioconductor.org/packages...ml/DESeq2.html

rajarapupriya 03-20-2014 05:05 AM

Awesome !! I appreciate your efforts !!

Thanks again
Priya


All times are GMT -8. The time now is 01:40 PM.

Powered by vBulletin® Version 3.8.9
Copyright ©2000 - 2019, vBulletin Solutions, Inc.