[BioC] EdgeR: problem with large logFC value

Gordon K Smyth smyth at wehi.EDU.AU
Fri May 25 02:22:21 CEST 2012


Dear Sandra,

It doesn't seem surprising to get large fold changes if the counts are all 
zero in one of groups (have you checked whether this is so?).

However you are using an old version of Bioconductor, and we only answer 
questions on the current version.  The current version of edgeR 
automatically moderates the logFC towards zero to avoid very large values, 
even when all the counts are zero in one of the groups.

Best wishes
Gordon

> Date: Wed, 23 May 2012 17:12:48 -0500 (CDT)
> From: sgonzalez at ira.cinvestav.mx
> To: bioconductor at r-project.org
> Subject: [BioC] EdgeR: problem with large logFC value
>
>
> Dear all,
>
> I am doing DGE multifactorial analysis with EdgeR but I have a problem
> with somes locus I obtained a large logFC value , perhaps because any
> comparision there is some zero.
> Is this possible? or maybe I have some error
>
> Thank for your help, waiting your response.
> Regards,
>
> This is the scrip that I am using:
>
> d.all<- readDGE(counts.all, skip = 5, comment.char = "!")
> colnames(d.all$counts)<-counts.all[,3]
> cpmd.all<-cpm(d.all)
> d.all <- d.all[rowSums(cpmd.all > 1) >= 2, ]
> d.all<- calcNormFactors(d.all)
> designall<-model.matrix(~time+time:d.all$samples$group)
> d.all <- estimateGLMCommonDisp(d.all, designall)
> glmfit.all <- glmFit(d.all, designall,dispersion = d.all$common.dispersion)
> lrt.all <- glmLRT(d.all, glmfit.all, coef=4)
> topTags(lrt.all,sort.by = "logFC")
> 	logConc      logFC       LR      P.Value          FDR
> AT1G22220 -12.30768 -144269488 62.70746 2.398056e-15 2.995171e-12
> AT1G08430 -13.45102 -144269487 37.11160 1.115581e-09 2.960891e-07
> AT4G11180 -13.84678  144269487 38.22337 6.309211e-10 1.835116e-07
> AT5G41880 -13.36909 -144269486 33.88640 5.842556e-09 1.255052e-06
> AT1G04610 -13.41465 -144269486 33.44700 7.323248e-09 1.480900e-06
> AT2G24850 -14.18267 -144269486 25.85859 3.673666e-07 3.979742e-05
> AT5G44635 -13.98784 -144269486 24.61663 6.994557e-07 6.736796e-05
> AT1G35690 -13.60305  144269486 27.92875 1.258656e-07 1.590776e-05
> AT1G67328 -13.27975 -144269486 24.43246 7.696091e-07 7.262716e-05
> AT1G51460 -13.77582 -144269486 24.12274 9.038647e-07 8.123359e-05
>
> sessionInfo()
> R version 2.14.1 (2011-12-22)
> Platform: x86_64-pc-linux-gnu (64-bit)
>
> locale:
> [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
> [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
> [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
> [7] LC_PAPER=C                 LC_NAME=C
> [9] LC_ADDRESS=C               LC_TELEPHONE=C
> [11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C
>
> attached base packages:
> [1] stats     graphics  grDevices utils     datasets  methods   base
>
> other attached packages:
> [1] edgeR_2.5.5  limma_3.10.3
>
> loaded via a namespace (and not attached):
> [1] tools_2.14.1
>
>
> Sandra Isabel Gonzalez Morales
> Graduet Student PhD
> Centro de Investigacion y de Estudios Avanzados del I.P.N
> Laboratorio Nacional deGenomica
> para la Biodiversidad
> Km. 9.6 Libramiento Norte Carretera Irapuato-Leon
> Irapuato, Gto. M?xico
>

______________________________________________________________________
The information in this email is confidential and intend...{{dropped:4}}



More information about the Bioconductor mailing list