Error in gammaFitEM for library(lumi) [Possible Bug]
2
0
Entering edit mode
@chang-jun-park-5086
Last seen 10.2 years ago
Dear Bioconductor developers, My name is Chris Park who is working as a bioinformatician in University Health Network in Toronto. I'm currently working with Illumina methylation data by preprocessing them in R, and discovered a problem in which it could potentially be a bug. The bug happens when I call function for Methylation calls: lumiMethyStatus {lumi}, which gives this error message: Error in if (truncate) { : argument is not interpretable as logical > traceback() 3: gammaFitEM(x, ...) 2: methylationCall(M[, i], ...) 1: lumiMethyStatus(data.lumiMethy.bc.adj.quantile[, c(1:8)]) As shown on traceback(), the error is in gammaFitEM. After looking at the code, I found that if iteration reaches the maxInteration (hard-coded as 50), it will enter: 1 if (iter == maxIteration) { 2 Mode <- c(s[1] + (k[1] - 1) * theta[1], s[2] - (k[2] - 3 1) * theta[2]) 4 f1 <- dgamma(x - s[1], shape = k[1], scale = theta[1]) 5 f2 <- dgamma(s[2] - x, shape = k[2], scale = theta[2]) 6 if (truncate) { 7 f1 <- f1/(pgamma(Mode[2] - s[1], shape = k[1], scale = theta[1], 8 lower.tail = TRUE)) 9 f2 <- f2/(pgamma(s[2] - Mode[1], shape = k[2], scale = theta[2], 10 lower.tail = TRUE)) 11 z1 <- p[1] * f1/(p[1] * f1 + p[2] * f2) 12 z1[x > Mode[2]] <- 0 13 z1[x < Mode[1]] <- 1 14 f1[x > Mode[2]] <- 0 15 f2[x < Mode[1]] <- 0 16 } 17 else { 18 z1 <- p[1] * f1/(p[1] * f1 + p[2] * f2) 19 } 20 z2 <- 1 - z1 21 } Where on line 6, it checks for logical argument: truncate, but this logical variable is nowhere to be found. What do you suggest is the best solution is? Theoretically, what is a reason to decide whether to "truncate" or not when we reach maximum iteration of the "while" loop? I am using the latest R version available with the most recent package, as shown in sessionInfo(): > sessionInfo() R version 2.14.1 (2011-12-22) Platform: x86_64-pc-mingw32/x64 (64-bit) locale: [1] LC_COLLATE=English_Canada.1252 LC_CTYPE=English_Canada.1252 [3] LC_MONETARY=English_Canada.1252 LC_NUMERIC=C [5] LC_TIME=English_Canada.1252 attached base packages: [1] stats graphics grDevices utils datasets methods base other attached packages: [1] KernSmooth_2.23-7 IlluminaHumanMethylation450k.db_1.4.6 [3] org.Hs.eg.db_2.6.4 RSQLite_0.11.1 [5] DBI_0.2-5 AnnotationDbi_1.16.11 [7] lumi_2.6.0 nleqslv_1.9.2 [9] methylumi_2.0.4 Biobase_2.14.0 loaded via a namespace (and not attached): [1] affy_1.32.1 affyio_1.22.0 annotate_1.32.1 [4] BiocInstaller_1.2.1 grid_2.14.1 hdrcde_2.15 [7] IRanges_1.12.5 lattice_0.20-0 MASS_7.3-16 [10] Matrix_1.0-3 mgcv_1.7-13 nlme_3.1-103 [13] preprocessCore_1.16.0 tools_2.14.1 xtable_1.6-0 [16] zlibbioc_1.0.0 [[alternative HTML version deleted]]
Preprocessing Network Preprocessing Network • 1.2k views
ADD COMMENT
0
Entering edit mode
@elodie-chapeaublanc-4525
Last seen 10.2 years ago
Hello, I have the same problem with my own Illumina Methylation 27k data. This error doesn't appear with example.lumiMeth object. Do you have any solution ? Thanks, Elodie
ADD COMMENT
0
Entering edit mode
Pan Du ▴ 440
@pan-du-4636
Last seen 10.2 years ago
Thanks! Elodie It's a bug after I updated the code last time. I have fixed it by adding an additional parameter "truncate=FALSE" to the gammaFitEM function. I will upload the change together with many other methylation related updates at the end of this week. Pan > Subject: Re: [BioC] Error in gammaFitEM for library(lumi) [Possible Bug] > Date: Tue, 14 Feb 2012 08:23:57 +0000 > From: Elodie Chapeaublanc <elodie.chapeaublanc at="" curie.fr=""> > To: <bioconductor at="" stat.math.ethz.ch=""> > > Hello, > My name is Elodie Chapeaublanc. I am a Bioinformatician in Institute CURIE at Paris. > I want to use lumi package in order to preprocess Illumina methylation 27K data. > When I try to use lumiMethyStatus on matrix data, an error appear : > Error in if (truncate) { : argument is not interpretable as logical > I haven't got error with example.lumiMethy object. > How can I resolve this error? > Thanks, > Elodie
ADD COMMENT

Login before adding your answer.

Traffic: 477 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6