[BioC] AgiMicroRna and Replicates

[Richard Friedman]

Dear Karthik,

On Jun 14, 2012, at 9:45 AM, Karthik K N wrote:

> Dear Rich,
>
> Thanks a lot for your reply. I am just worried because somewhere I  
> remember reading that LIMMA can't be used without replicates ( in  
> case of mRNA microarray analysis); so just wanted to check if it is  
> the same with AgiMicroRna package also.

I think that LIMMA can be used without replicates but it cannot give  
you a p-value. I suggest
a minimum of 3 biological replicates per sample.

>
> Also, Do you think it is a good idea to  carry out GeneSpring  
> analysis upon the data analyzed by biocondcutor? Will it give any  
> statistically more reliable output that using either of them alone?

I do not know how GeneSpring normalizes Agilent MicroRNA  data but  
AgiMicroRna  is the best way to
normalize such data described in the open literature of which I am  
aware.
The statistical methods in LIMMA are more reliable than those in  
GeneSpring in general. However
with only one replicate of each condition neither program can give a p- 
value, only a log2FC.
You can get the same result in an excel spreadsheet,


>
> Also, since the data from bioconductor has already normalized, if we  
> again give this as an input file for genespring and go on with  
> analysis, then do we need to do normalization step again in  
> GeneSpirng? Won't this be a double-normalization?

I haven't used GeneSpring for a long time. I am not sure what it will  
do to your data.
If you only have one replicate you might  consider using AgiMicroRNA  
to normalize and Excel to
compute the log2FC.

Best  wishes,
Rich



>
> I am new to R/Bioconductor, so any suggestions from you will be  
> extremely helpful.
>
> Thanks a lot,
>
> Karthik
>
>
> On Thu, Jun 14, 2012 at 7:08 PM, Richard Friedman <friedman at cancercenter.columbia.edu 
> > wrote:
> Dear Karthik,
>
>        I am pretty sure that AgiMicroRna will normalize one treated  
> and
> one control. The problem comes later in terms of the reproducibility
> of the effect, or to phrase it differently, whether the observed  
> effect
> is a general statement  about the population of treatments and  
> controls.
> In more specific terms, the subsequent LIMMA analysis will not
> compute a p-value.
>
> With hopes that this helps,
> Rich
> ------------------------------------------------------------
> Richard A. Friedman, PhD
> Associate Research Scientist,
> Biomedical Informatics Shared Resource
> Herbert Irving Comprehensive Cancer Center (HICCC)
> Lecturer,
> Department of Biomedical Informatics (DBMI)
> Educational Coordinator,
> Center for Computational Biology and Bioinformatics (C2B2)/
> National Center for Multiscale Analysis of Genomic Networks (MAGNet)
> Room 824
> Irving Cancer Research Center
> Columbia University
> 1130 St. Nicholas Ave
> New York, NY 10032
> (212)851-4765 (voice)
> friedman at cancercenter.columbia.edu
> http://cancercenter.columbia.edu/~friedman/
>
> "School is an evil plot to suppress my individuality"
>
> Rose Friedman, age15
>
>
>
>
>
>
>
>
>
>
>
> On Jun 14, 2012, at 5:51 AM, Karthik K N wrote:
>
> Dear Members,
>
> Do we need replicates to carry out analysis with AgiMicroRna package  
> in
> bioconductor? I have one control and one treated samples. Can I go  
> ahead
> with AgiMicroRna with these two datasets?
>
> Thank you.
>
> -- 
> Karthik K.N
>
>        [[alternative HTML version deleted]]
>
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at r-project.org
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor
>
>
>
>
> -- 
> Karthik K.N
> Cancer Discovery Biology Laboratory
> Division of Molecular Medicine
> Amrita Center for Nanosciences and Molecular Medicine
> Amrita Institute of Medical Sciences
> AIMS-Ponekkara (P.O), Kochi
> Cochin, Kerala - 682 041
> +91-484-280 1234 x 8720
> +91-9400193907
>