Despite matching the results from GCOS, I have never seen a mean that low or a negative value...so I assumed it was weird. Furthermore, I used affyQCReport which shows mean Log2 intensity on all arrays of 5 and an IQR between 4 and 12, which is what I would expect from the RMA.
RMA Weird Distribution
I'm trying to normalize GSE27657 with RMA in R. However, I am getting a weird distribution of probeset expressions with mean=2.16, max=12.77, and min=-0.3.
QC seemed fine on these chips to my eye, and the QC report showed a different (What I would expect) distribution of expression values.
I just used justRMA(). Furthermore, frma shows normal values. Any ideas?
> sessionInfo()
R version 3.1.1 (2014-07-10)
Platform: x86_64-pc-linux-gnu (64-bit)
locale:
[1] LC_CTYPE=en_CA.UTF-8 LC_NUMERIC=C LC_TIME=en_CA.UTF-8
[4] LC_COLLATE=en_CA.UTF-8 LC_MONETARY=en_CA.UTF-8 LC_MESSAGES=en_CA.UTF-8
[7] LC_PAPER=en_CA.UTF-8 LC_NAME=C LC_ADDRESS=C
[10] LC_TELEPHONE=C LC_MEASUREMENT=en_CA.UTF-8 LC_IDENTIFICATION=C
attached base packages:
[1] parallel stats graphics grDevices utils datasets methods base
other attached packages:
[1] affy_1.44.0 Biobase_2.26.0 BiocGenerics_0.12.0
loaded via a namespace (and not attached):
[1] affyio_1.34.0 BiocInstaller_1.16.0 preprocessCore_1.28.0 tools_3.1.1
[5] zlibbioc_1.12.0
1 answer
Why do you think the distribution is "weird"? The numbers match those generated by the GCOS RMA normalisation used by the authors, which you can access using GEOquery:
library(GEOquery)
gse <- getGEO("GSE27657")
gse <- gse$GSE27657_series_matrix.txt.gz
e <- exprs(gse)
fivenum(e)
[1] -0.306705 0.672482 1.452132 3.198111 12.773470
You should also clarify what QC you used, why you think it is as "expected" and why you think the frma values are "normal".
Log in to answer this question.
More posts like this
-
bsseq: Failing to reproduce the results
written by Mithil Gaikwad •
-
[R/Rstudio] Can't load installed packages with "no title available" message
written by Min Dai
-
KEGGREST -- gene list
written by haiying.kong
-
Semantic Similarity against all MeSH categories
written by rzjfr •
-
Error running getBamCounts from ExomeDepth
written by alsanju •
-
MLSeq package error
written by jonessara770
-
Question: Getgeo Error - line 1 did not have 6 elements
written by jimmy_zeng
-
Missing ensembl_ids in biomaRt uniprot query
written by A. Domingues
-
Cannot coerce class 'structure("IRanges", package = "IRanges")' into a data.frame error.
written by Aishwarya.Kulkarni •
-
Error in .normargSubject(subject) : active masks are not supported yet, please complain! in match…
written by Aishwarya.Kulkarni •
could you check if its the case of RNA degradation, do you also have an access to cel files?
Ran it. Not sure how to interprety results though. Running the summary function on the first function (AffyRNAdeg) results in very small p values (max=5e-11), and slopes of mean 3.85.
On all of the samples just run;
see that how the plot looks like, can u share the plot?
data looks comparable,
I think reinstall affy in your R, start analyzing with .cel files (you already have scripts so should not take much time), go for gcrma or quantile normalization, and see if it works this time
Alright. I suppose I will compare RMA vs gcRMA vs fRMA to see if there are any differences.