[BioC] The issue with Report Pearson instead of Spearman

Wolfgang Huber whuber at embl.de
Tue Aug 17 23:38:49 CEST 2010


Dear Ning

you are right, the option corr.method="spearman" is currently hard-coded 
into the writeReport function of the cellHTS2 package. One opinion on 
that is that I have rarely seen cases where these two methods produce 
substantially different results, and if they do, I would probably want 
to study the data very carefully (outside an automated workflow like 
writeReport) to understand why.

Still, if you really need to use that different option, the solution is 
to make your own copy of the writeReport function
(https://hedgehog.fhcrc.org/bioconductor/trunk/madman/Rpacks/cellHTS2/R/writeReport.R 
   user: readonly  password: readonly) and edit it to your liking. 
That's one of the beauties of open source :)

You could also try to convince the package maintainer to add a relevant 
argument to that (already monstrous) function.	

	Best wishes
	Wolfgang

On Aug/17/10 7:40 PM, Ning wrote:
> Hi Wolfgang,
>
> Thank you for your reply. Actually my question is how I can let the results of
> Pearson correlation showed in the report when I use the function of writeReport.
> As every time I use this function, only the results of Spearman rank correlation
> were showed in the report, even I have called
>
> repAgree<- getMeasureRepAgreement(xn, corr.method = "pearson")
>
> I checked the instruction of writeReport function, but cannot find the way to
> put the “repAgree” in it.
>
> My code of writeReport is
>
> setSettings(list(plateList=list(reproducibility=list(include=TRUE, map=TRUE),
> intensities=list(include=TRUE, range=c(-0.5,1), map=TRUE)),
> screenSummary=list(scores=list(range=c(-4, 8), map=TRUE))))
> out<- writeReport(raw=x, normalized=xn, scored=xsc,map=TRUE, force=TRUE)
>
> Do you have any idea about where I put the “repAgree” in?
> Many thanks!
>
> Ning
>
> The following is my script
>
> # read platelist
> x<- readPlateList("Platelist.txt",
>                      name=experimentName,
>                      path=dataPath)
> #Configure X
> x<- configure(x,
>                 descripFile="Description.txt",
>                 confFile="Plateconf.txt",
>                 logFile="Screenlog.txt",
>                 path=dataPath)
> #Normalize X
>   xn<- normalizePlates(x,
>         scale="multiplicative",
>         log=FALSE,
>         method="NPI",
>         varianceAdjust="none")
>
> #Calculate the Pearson correlation with normalized “xn”
> repAgree<- getMeasureRepAgreement(xn, corr.method = "pearson")
>
> #Scored “xn”
> xsc<- scoreReplicates(xn, sign="-", method="zscore")
>
> # summarize
> xsc<- summarizeReplicates(xsc, summary="median")
>
> # annotate
> xsc<- annotate(xn, geneIDFile="GeneIDs.txt",  path=dataPath)
>
> # write report
> setSettings(list(plateList=list(reproducibility=list(include=TRUE, map=TRUE),
>              intensities=list(include=TRUE, range=c(-0.5,1), map=TRUE)),
>              screenSummary=list(scores=list(range=c(-4, 8), map=TRUE))))
> out<- writeReport(raw=x, normalized=xn, scored=xsc,map=TRUE, force=TRUE)
>
> _______________________________________________
> 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

-- 


Wolfgang Huber
EMBL
http://www.embl.de/research/units/genome_biology/huber



More information about the Bioconductor mailing list