[BioC] Most stable gene pairs in array experiment

[David martin]

Exactly that what i want to do.
How could i do that on the whole matrix taking all pairs (or at least 
one fixed gene ,e.g geneA and compute all var comparisons 
var(geneA-geneB) var(geneA-geneC) var(geneA-geneC) and see which 
variance is better ??
thanks

Naomi Altman wrote:
> It sounds to me like you are looking for var(geneA-geneB).
> --Naomi
> 
> At 05:54 AM 10/19/2009, anna freni sterrantino wrote:
>> Hi  David,
>> not sure what do you mean with stable,
>> but you might be interested in correlation,
>>
>> a=matrix(sample(1:100),4,5)
>> rownames(a)=paste("gene", letters[1:4])
>> colnames(a)=paste("cond", letters[1:5])
>>   >a
>> cond a cond b cond c cond d cond e
>> gene a     95     31      3      9     93
>> gene b     16     67     83     81     86
>> gene c     59     79     44     77     39
>> gene d     36     92     41     57     66
>> > cor(t(a))
>>           gene a     gene b     gene c     gene d
>> gene a  1.0000000 -0.5362894 -0.3830295 -0.1109239
>> gene b -0.5362894  1.0000000 -0.1710537  0.3790986
>> gene c -0.3830295 -0.1710537  1.0000000  0.4612277
>> gene d -0.1109239  0.3790986  0.4612277  1.0000000
>>
>> and then  across all the pairs the most correlated will be those
>> that have  a correlation value that is close to  |1|.
>> The correlation tells you how much close are two variables in terms
>> of linear relationship.
>>
>> Hope this helps.
>> Cheers
>>
>> A
>>
>>
>>
>> Anna Freni Sterrantino
>> Ph.D Student
>> Department of Statistics
>> University of Bologna, Italy
>> via Belle Arti 41, 40124 BO.
>>
>>
>>
>>
>> ________________________________
>> Da: David martin <vilanew at gmail.com>
>> A: bioconductor at stat.math.ethz.ch
>> Inviato: Lun 19 ottobre 2009, 10:37:08
>> Oggetto: [BioC] Most stable gene pairs in array experiment
>>
>> Hi,
>> I have the following matrix with normalized log2 values:
>> CondA    CondB    CondC    CondD    CondE
>> geneA    -6.19    -5.74    -5.82    -5    -5.59
>> geneB    -6.33    -5.32    -5.6    -4.88    -5.39
>> geneC    -6.15    -6.07    -5.6    -4.88    -5.9
>> geneD    -6.57    -6.11    -6.36    -5.36    -5.96
>> geneD    -6.74    -6.2    -5.49    -5.35    -5.95
>> geneE    -6.75    -6.24    -5.73    -5.63    -6.02
>>
>>
>> Created as follows:
>> geneA<-c(-6.19,   -5.74,   -5.82,   -5,  -5.59)
>> geneB<-c(-6.33,   -5.32,   -5.6,    -4.88,   -5.39)
>> geneC<-c(-6.15, -6.07, -5.6, -4.88, -5.9)
>> geneD<-c(-6.57,   -6.11,   -6.36,   -5.36,   -5.96)
>> geneD<-c(-6.74,   -6.2,    -5.49,   -5.35,   -5.95)
>> geneE<-c(-6.75,   -6.24,   -5.73,   -5.63,   -6.02)
>>
>> mygenes<-rbind(geneA, geneB, geneC, geneD, geneE)
>> colnames(mygenes)<-c("CondA",   "CondB",   "CondC",   "CondD",
>> "CondE")
>>
>> I'm looking for most stable pair genes across conditions. I'm not 
>> looking for individual gene variance but really for most stable pairs 
>> ratios.
>> For e.g What is the variance of geneA vs geneB across all conditions. 
>> What is the most stable pair ?
>>
>> Any help would be appreciated.
>>
>> david
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at stat.math.ethz.ch
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>> Search the archives: 
>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>
>>
>>
>>
>>         [[alternative HTML version deleted]]
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at stat.math.ethz.ch
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>> Search the archives: 
>> http://news.gmane.org/gmane.science.biology.informatics.conductor
> 
> Naomi S. Altman                                814-865-3791 (voice)
> Associate Professor
> Dept. of Statistics                              814-863-7114 (fax)
> Penn State University                         814-865-1348 (Statistics)
> University Park, PA 16802-2111
> 
> _______________________________________________
> Bioconductor mailing list
> Bioconductor at stat.math.ethz.ch
> https://stat.ethz.ch/mailman/listinfo/bioconductor
> Search the archives: 
> http://news.gmane.org/gmane.science.biology.informatics.conductor
>