[BioC] Filtering SAM results

[Holger Schwender]

Hi Assa,

take a look at findDelta. This function does exactly what you would like to have -- but only for FDR or Called. For False, you have to hack siggenes:::findNumber.

Best,
Holger



-------- Original-Nachricht --------
> Datum: Wed, 25 Mar 2009 12:26:11 +0100
> Von: Assa Yeroslaviz <frymor at gmail.com>
> An: bioconductor at stat.math.ethz.ch
> Betreff: [BioC] Filtering SAM results

> Hi everybody,
> 
> I would like to automate my sam output files.
> 
> I'm trying to do many SAM analysis together and would like to plot the
> results and export the list of sig. genes to a file. The point is I would
> like to try and make it with a R Script, so that I don't need to go into
> it
> and change the parameter every time.
> 
> my sam command is:
> 
> > samresult    <- sam(groups, cl=classes, var.equal=FALSE, B=100,
> include.zero=FALSE, gene.names=gene_names, q.version=2, R.fold=0,
> na.replace=FALSE, rand=123)
>  and I have two thresholds I would like to look into:
> 1. false positives  - the first threshold is that false < 1 (here number 5
> with 0.986 false)
>    Delta    p0 False Called     FDR
> 1    8.0 0.161 1.143     79 0.00233
> 2    8.1 0.161 1.071     74 0.00233
> 3    8.2 0.161 1.029     71 0.00233
> 4    8.3 0.161 1.000     69 0.00233
> 5    8.4 0.161 0.986     68 0.00233
> 6    8.5 0.161 0.971     67 0.00233
> 7    8.6 0.161 0.914     63 0.00233
> 8    8.7 0.161 0.914     63 0.00233
> 9    8.8 0.161 0.871     61 0.00230
> 10   8.9 0.161 0.829     58 0.00230
> 11   9.0 0.161 0.814     57 0.00230
> 12   9.1 0.161 0.800     56 0.00230
> 13   9.2 0.161 0.771     54 0.00230
> 14   9.3 0.161 0.771     54 0.00230
> 15   9.4 0.161 0.743     52 0.00230
> 16   9.5 0.161 0.729     51 0.00230
> 17   9.6 0.161 0.671     47 0.00230
> 18   9.7 0.161 0.657     46 0.00230
> 19   9.8 0.161 0.643     45 0.00230
> 20   9.9 0.161 0.629     44 0.00230
> 21  10.0 0.161 0.557     39 0.00230
> 
> The second parameter will probably be the FDR. But I still didn't find the
> right %FDR to suit to my analysis.
> 
> the output I'm using till now is:
> > print(samresult1h, seq(8,10,0.1))
> > plot(samresult1h, 8.4)
> 
> But as you can imagine it is a tedious work to do it for many files.
> 
> My question is - Is there a possibility to tell R that I want to look only
> for the first delta value where False is smaller than 1 or the FDR value
> has
> reached a certain %?
> 
> 
> THX for any help
> 
> Assa
> 
> 	[[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

--