Problems importing Illumina raw data with lumi R package
2
0
Entering edit mode
svlachavas ▴ 800
@svlachavas-7225
Last seen 3 days ago
Germany/Heidelberg/German Cancer Resear…

Dear Bioconductor Community,

i recently aquired 3 txt files from a microarray experiment including Illumina Human  HT-12 v4 beadchip and i used the lumi package and vignette to preprossess the data in the following way

rawdata <- "SampleProbeProfile_Dimitra3.txt" # includes columns refering to PROBE_ID, AVG_Signal,..
smplfile <- "SampleTableControl_Dimitra3.txt # phenoData information about the samples and the Groups
ctrlprobes <- "ControlProbeProfile_Dimitra3.txt" # information about the control probes
lib_mapping <- "lumiHumanIDMapping"
rawdata.lumi <- lumiR.batch(rawdata, lib.mapping=lib_mapping, sampleInfoFile= smplfile)

Inputting the data ...

Adding nuID to the data ...

Attaching package: ‘IRanges’

The following object is masked from ‘package:simpleaffy’:

members

No Quality Control assessment of the object because it is not a "LumiBatch" object.

Inputting the data ...
Error in checkAtAssignment("ExpressionSet", "controlData", "data.frame") :
‘controlData’ is not a slot in class “ExpressionSet”

class(rawdata.lumi)
[1] "ExpressionSet"
attr(,"package")
[1] "Biobase"

also i tried lumi.T <- lumiT(rawdata.lumi) but
Error in lumiT(rawdata.lumi) : Slot se.exprs is required!

Any ideas or help ??

Furthermore, i have found from limma a way of preprocessing illumina raw data, but i dont know which method (including normalization and background correction) is more effective and superior for illumina data

Illumina Human HT-12 v4 lumi limma microarray • 2.7k views
0
Entering edit mode

Hi user svlachavas

I am new to this analysis and would like your valuable suggestions to process my data. If its fine for you, kindly contact me on mathew.mano@gmail.com

1
Entering edit mode
Pan Du ▴ 80
@pan-du-4535
Last seen 7.7 years ago
United States
0
Entering edit mode

Dear Pan Du,

as im a beginner with Bioconductor experience from the last 8 months, so far i have analyzed only affymetrix microarray data and for Illumina  was the first time to analyze, thats why i also asked about the other alternative possibility of preprocessing with limma. I searched also other similar posts before posting but i didnt found your above explanation, for which i would like to thank you because i checked my "SampleProbeProfile_Dimitra3.txt" and the columns STDEV or STDERR do not exist, only the above i mentioned.

0
Entering edit mode

One more important question i would like to ask, is that even i miss the above columns required especially for lumi, could i still preprocess the data with limma and read.ilmn and neqc functions ?

0
Entering edit mode

Hi,

I have the same problem like yours. My raw data file has only "Probe ID" and "AVG Signal" columns. I started using Lumi package for data processing and Normalization.

file = 'BCCSED.txt'
x.lumi = lumiR.batch(file, lib.mapping = 'lumiHumanIDMapping')

Inputting the data ...

Adding nuID to the data ...
No Quality Control assessment of the object because it is not a "LumiBatch" object.

Could you please help me how to do now? Along with the raw data file I have another two files like "Probe annotation file" and "Sample Annotation file"

1
Entering edit mode
@gordon-smyth
Last seen 1 hour ago
WEHI, Melbourne, Australia

"Furthermore, i have found from limma a way of preprocessing illumina raw data, but i dont know which method (including normalization and background correction) is more effective and superior for illumina data."

The limma User's Guide gives a complete worked case study of reading and processing Illumina Beadchips in Section 17.3. You can read and normalize the Beadchip data even if your files are missing the STDEV or STDERR columns.

For a comparison of normalization and background methods for Illumina BeadChips, see:

0
Entering edit mode

Dear Mr. Gordon Smyth,

thank you for your notifications. I found the guide to preprossess Illumina with limma package, and i also have two naive but very important questions:

1. In order to perform some quality plots both before and after normalization, could i use the output of read,ilmn (i.e. x <-read.ilmn) and then extract the expression of x: raw.matrix <- x$E and after normalization norm.matrix <- y$E(where y the output of neqc) to make some plots, i.e: boxplot(raw.matrix), boxplot(norm.matrix) and to use some packages like arrayQualityMetrics ?

2. Secondly, regarding the important issue of annotation, in lumi with the use of the function

y <- addNuID2lumi(rawdata.lumi, lib.mapping='lumiHumanIDMapping') the names of the probes are mapped to nuIDs and then these are used to get the gene symbols and other annotation sources.

In limma could i perform something similar, or directly, with the function select() and  "lumiHumanAll.db" ? And overall, is the annotation consistent between the two packages ?

Last but not least, i checked your above paper and also i tried to preprossess the data with read.ilmn

My initial file of the summary probe profile intensities looks like this :

PROBE_ID         SYMBOL    1A.AVG_Signal    1A.Detection Pval ....

ILMN_1762337    7A5            113.6922              0.4623377.....

and after the use of read.imn:

1A           1B         2A           2B            3A           3B          4A          4B          5A
ILMN_1762337 113.6922 104.2509 104.3843 113.8150 111.22100 102.4746 113.6853 121.3207 118.4310
ILMN_2055271 162.7333 119.7672 126.7262 144.6663 122.19350 119.3691 143.7316 134.9968 125.3691
ILMN_1736007 114.4497 110.0415 118.1159 117.1557  97.70956 104.0438 115.2251 135.1003 107.3504
ILMN_2383229 122.7193 103.4715 104.9192 114.8986 101.92230 104.3109 121.4065 109.3550 108.3537
ILMN_1806310 134.9052 118.6454 103.8566 110.8623 110.44460 124.6512 127.5805 117.3484 114.1275
ILMN_1779670 130.3925 128.4179 111.4081 117.3661 111.56040 105.9566 123.0708 126.5529 110.3249

My last question is that (and please correct me if i have misunderstood it) the above intensities(from the txt file i got and import with limma) it seems that they have already been background corrected, could anyway i use neqc or because neqc also performs a background correction is inappropriate ?

Thank you for your consideration on this matter !!

0
Entering edit mode

If you want help with annotation and background correction using the limma package, could you please post a question about that? The current question is about reading data using lumi.