How to compute exression ratio of Human Exon 1.0 ST array on probe level

[tujchl]

Hi guys:
Recently I search some chip data (Human Exon 1.0 ST array) from GEO, I found the annotation file on GEO is based on either probeset level (usually 4 probes) or even transcripts level. My question is how can I get expression value on probe level. I search on Bioconductor and want to find some packages. Unfortunately I don`t find any package suitable for me.
could you please recommend some packages or solutions for me?

thank you in advance

[malachig]

Getting probe-level data out of Affymetrix exon arrays

If you have the original .CEL values, one option it to get RAW probe values out of these using 'Affymetrix Power Tools'. This software is a command line tool that works in windows (I know crazy...). It does work though. In particular the 'apt-cel-extract' function should allow you to get the individual probe level values. Probe level annotations are available here: GeneChip Human Exon 1.0 ST Array

[malachig]

If you just want to re-process the data to the exon or gene level (including using your own custom annotation) then there are several options including: RMA Express, Expression Console, affy/BioConductor, Affymetrix Power Tools and others.

[kenosaki]

What data set did you download from? Tell us GEO accession # and downloaded dataset. It may help us to tell you how to get expression value.

[tujchl]

Thank you malachig! Thank you kenosaki!
the GEO accession NO. is GSE12941. Its annotation file is GPL5175, you can see the results directly download from GEO is transcripts based. I want to see the ratio based on probe level.

[kenosaki]

Ok, you can get CEL file. I usually download the CEL file when I want to probe level signals.
I like Expression Console to analyze it.

You need registration (free) to get the Expression Console. Once you installed it, you download library file and annotation file. Analysis will be run either Core, Extend, or Full. Choose Core, because Core probes are linked to the validated genes. Then RMA or whatever you like anlaysis will be done on probe level.
Export probe level result with certain annotations.
And also run gene level analysis and export gene level result.
Be carefull, "transcript ID of probe level result" is equal to "probe ID of gene level result".
Probe level results and Gene level results are now be able to be combined.

Well, this may not be the best short-cut for your purpose.
Take care.

[tujchl]

Thank you kenosaki, you help me a lot. I have dowloaded Expression Console and Now installing library, I will try it, thank you again!
One more question~~
Exon array could be analyzed on probeset, exon and transcripts level respectively. To my knowledge each probeset , exon and transcripts DO have their own ID which can be serached on NetAffy.
My question is whether the ID is unique? (I mean whether ID could overlappe among probeset , exon and transcripts ID) and don`t change when version update (I mean one id could not point to different probe seq in different chip version).
In the GSE12941, Its probeset ID such as 3553690, could be transcript cluster ID NOT probeset ID when I search on NetAffy. could you please tell me why?

Finally, Thank you again

[tujchl]

One more question, kenosaki
when I load CEL file and Expression Console tell me to download "HuEx-1_0-st-v1.cdf" file (for Expression arrays) or "HuEx-1_0-st-v1.exon_analysis_configuration" (for Exon arrays). I just can find v2 on Affy, could you please help me again?

[tujchl]

malachig, I use the package RMAExpress, and do you know how to export expression value on probe level? Thank you

[kenosaki]

Question 1,
A: Probeset ID is unique, as long as of the same chip. I don't think there is a change on different version. Every probe was designed from a transcript cluster (RefSeq RNA, EST,etc), so this cluster ID should be decided at the time of array design, it means they don't change at all. You better to double check to Affy.

For Question 2,
Yes I see the same error.
You have "HuEx-1_0-st-v2.exon_analysis_configuration" file in your library folder, and you need "HuEx-1_0-st-v1.exon_analysis_configuration" insted.
This file is not officially supported, so you need to ask Affy to get this file.
This is small 6kb file. I can send you by e-mail if you want.

[tujchl]

It is very kinds of you to offer me the annotation file-_-, My e-mail is delete[/email].
a little question about Expression Console:
After I load the CEL file, there are only two categories of available analysis methods: Gene Level and Exon Level (I choose exon level core), whithout probe level. After analysis, I can only export the results based on probesets....
could you please tell me more about operation?

Thank you

[kenosaki]

I sent the conf file to you by e-mail. Locate this file into the libray folder of expression console
I suggest that you delete your e-mail address from this thread now.

By the way, exon level means probe level, as you expected.
Operation? Depends on what you want to do. Could you look at Help or tutorial that Affy provided? It helps the operation pretty much.
I know they have other technical documents about exon array.
And I believe they support you.

If you can't find any help sitll, feel freet to ask me

[tujchl]

OK, I have the the e-mail, Thank you very very much

[honey]

I was wondering if from the exon array can I get exon level expression analysis or transcript level. I am interested in more of exon level expression.
Has anyone successfully completed such analysis with exon arrays.

Thanks

[gargvin2003]

Hi kenosaki,
Could you please send me the "HuEx-1_0-st-v1.exon_analysis_configuration" file. I am having the same problem with opening the "CEL" file in expression console.

[email protected]