Hi everyone,
Please click the following URL for RNA degradation plot.
http://i.imgur.com/IacdSwX.png
Could you please help me how to interpret the graph?
Thank you
There is no interpretation to be done. You are using a method intended for 3'-biased arrays with a random-primer array. IMO, these plots were of limited use for the 3'-biased arrays, and are of no use for random primer arrays.
Dear Venkatesh,
As I have mentioned in my last reply, see:
A: Error: Number of PMs or MMs is zero (xps package)
RNA degradation plots were used for Affymetrix IVT-arrays and not for arrays using random primers.
I have also said:
'Thus, in principle for whole genome/exon arrays RNA degradation plots are probable useless'.
However, after some further thinking and remembering my own tests I can at least say the following:
Although the curves do look different than the curves expected for IVT-arrays, my experience is
that the curves within an experiment should at least show the same behavior, which is the case
for your curves.
I hope that Jim can agree with me. However, as Jim mentioned in an older reply, see:
A: RNA degradation plot: why curves are so smooth?
'I think you will find that plotting the densities and looking at PCA plots will be more informative
than the RNA degradation plots.'
Please read Chapter 5 (especially 5.2) of vignette 'xps.pdf' for further ways to do quality control.
Regards,
Christian
The standard package for differential gene expression analysis is the Bioconductor package 'limma'.
The limma vignette has an extensive chapter for Affymetrix chips. Furthermore, numerous questions
how to use limma have been asked and answered on the support.bioconductor site.
For limma you can use the normalized data either as text file or you can convert the data to class
'ExpressionSet' as shown in Appendix A3 of vignette 'xps.pdf'.
Regards,
Christian
Hi Christian,
scheme.exon <-import.exon.scheme("scheme_root",filedir=scmdir,layoutfile=paste(libdir,"*.clf",sep="/"),schemefile=paste(libdir,"*.pgf",sep="/"),probeset=paste(anndir,"*.probeset.csv",sep="/"),transcript=paste(anndir,"*.transcript.csv",sep="/"),verbose=TRUE)
Error in import.exon.scheme("scheme_root", filedir = scmdir, layoutfile = paste(libdir, :
missing CEL Layout File *.CLF
Is it possible to import the files without giving the name like the above?
HI Christian,
As you said I have used Limma for normalized values. I got a heatmap with all the genes and its normalized values. But how to get a heatmap only for Differential expressed genes?
By using prefilter and unifiltr I got diffeential expressed genes with p-value and fold change but it doesn't have normalized values. So, Is it possible to get heatmap for differential expressed genes ? Looking forward to your response.
Thank you !!
Dear Venkatesh,
As I have already mentioned in my last reply, please start a NEW thread for these type of
questions since these have nothing to do with RNA degradation plots!!!
1, What do you mean with 'normalized values'?
2, When you use limma, you do not need to use prefilter and unifilter from xps.
3, I suggest to look at function 'topTable()' from limma.
Regards,
Christian
No!
How should the program decide which files to import? Furthermore, all flexibility would be lost.
Regards,
Christian
P.S.:
For questions not related to the title please start a new thread.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Thanks for the reply. Preprocess with RMA is done with xps package. And now how to assess differential gene expression?