[BioC] incoherence with LPE

[Jain, Nitin]

Hello Sabrina,

The function "fdr.adjust" with "resamp" option gives the matrix showing the
cutoff values of x-critical at different target fdrs.
It is investigator's decision to choose the cutoff fdr - the function
fdr.adjust will give you a "guideline" corresponding to the chosen fdr.
 
You should select the genes based on the lpe result whose z-value is higher
than z-critical (at your chosen fdr).
 
For fdr computation, the null distribution is derived from random sampling
within each bin (refer to my recent paper (Jul 05) in BMC-Bioinformatics).
 
Just FYI, I updated the LPE package (version 1.3.0) at bioconductor some
time ago - it is still in the Bioc-devel - changes will show in the released
version next month with the release of new Bioc version.

Have attached the new version here.

Best,
Nitin
 
______________________
Nitin Jain, PhD
<nitin.jain at pfizer.com>
Non Clinical Statistics
Pfizer, Inc. (Groton, CT)
Bldg: 260, # 1451
Ph:  (860) 686-2526 (Office)
Fax: (860) 686-6170



-----Original Message-----
From: bioconductor-bounces at stat.math.ethz.ch
[mailto:bioconductor-bounces at stat.math.ethz.ch]On Behalf Of Sabrina
Carpentier
Sent: Tuesday, August 30, 2005 10:13 AM
To: Bioconductor at stat.math.ethz.ch
Subject: [BioC] incoherence with LPE


Dear users,

I use the LPE package to detect differentially expressed genes but I'm
pretty surprised that I didn't find the same genes according the order of my
groups

>library(LPE)
>ind_g1<-c(2,4:6)
>ind_g2<-c(1,3,7)
>group1<-data[,g1]
>group2<-data[,g2]
# Finding the baseline distribution of condition 1 and 2.
>var.1 <- baseOlig.error(group1, q=0.01)
>var.2 <- baseOlig.error(group2, q=0.01)
>lpe.result <- lpe(group1,group2, var.1,
var.2,probe.set.name=rownames(data))
>finalresult<-fdr.adjust(lpe.result,adjp="resamp")
>finalresult
      target.fdr z.critical
 [1,]      0.001   5.606208
 [2,]      0.010   5.606208
 [3,]      0.020   5.606208
 [4,]      0.030   5.606208
 [5,]      0.040   4.720516
 [6,]      0.050   4.123739
 [7,]      0.060   2.823956
 [8,]      0.070   2.593902
 [9,]      0.080   2.534597
[10,]      0.090   2.482537
[11,]      0.100   2.375929
[12,]      0.150   2.015756
[13,]      0.200   1.630128
[14,]      0.500   1.500434

>lpe.inv<-lpe(group2,group1,var.2,var.1,probe.set.name=rownames(data))
>final.inv<- fdr.adjust(lpe.inv, adjp="resamp")
>final.inv
      target.fdr z.critical
 [1,]      0.001   3.096275
 [2,]      0.010   2.793027
 [3,]      0.020   2.593902
 [4,]      0.030   2.482537
 [5,]      0.040   2.482537
 [6,]      0.050   2.368713
 [7,]      0.060   2.368713
 [8,]      0.070   2.216946
 [9,]      0.080   2.091183
[10,]      0.090   1.774938
[11,]      0.100   1.639910
[12,]      0.150   1.547681
[13,]      0.200   1.537522
[14,]      0.500   1.459746

When I change the order of my groups, I didn't find the same z-critical. Do
you think that's right?

Thanks

Sabrina


Sabrina Carpentier
Service Bioinformatique
Institut Curie - Bat. Trouillet Rossignol (4e étage)
26 rue d'Ulm - 75248 Paris Cedex 5 - FRANCE
Tel : +33 1 42 34 65 21 
	[[alternative HTML version deleted]]




LEGAL NOTICE
Unless expressly stated otherwise, this message is confidential and may be privileged. It is intended for the addressee(s) only. Access to this E-mail by anyone else is unauthorized. If you are not an addressee, any disclosure or copying of the contents of this E-mail or any action taken (or not taken) in reliance on it is unauthorized and may be unlawful. If you are not an addressee, please inform the sender immediately.