Hi,

I'm analyzing some Human Exon Arrays from Affymetrix, from a pilot study, with oligo package, at transcript (core) level.

For the samples analyzed, I have some phenoData, as the expression levels (low / high, obtained by qRT-PCR) and mutational status of some genes. As it's a pilot study with very few number of chips, and unbalanced group sizes, I was asked to check whether these genes whose expression levels where known to be different, were actually differentialy expressed in the arrays.

The problem is that one of the genes I had to check has no annotation in pd.huex.1.0.st.v2 (I'm using pd.huex.1.0.st.v2_3.12.0). I checked for this gene in NetAffx Center, in order to know whether it was not represented in the array, or it has not been annotated, and I've found that there is a transcript cluster ID associated with the gene, and there's annotation. So, I've downloaded the last Affymetrix annotation at transcript level file (HuEx-1_0-st-v2.na35.hg19.transcript file) and the transcript cluster is now annotated, althought it wasn't there at least in the one I used when the arrays were firstly analyzed, with Affymetrix TAC Software, HuEx-1_0-st-v2 Transcript Cluster Annotations, CSV, Release 33.1.

So, I have two questions:

1) Is it possible for me to generate a pd.huex.1.0st annotation file based on the last annotation files from Affymtrix that could be integrated in the oligo workflow for analyzing these chips? how can oligo be said to use this file, instead of the one generated by Mr. Benilton Carvalho?

2) Is it possible to know the version of affymetrix annotation files used to generate pd.huex.1.0.st.v2_3.12.0?

Thank you in advance

Best regards,

Elena

The pd.huex.1.0.st.v2 package isn't used for annotation. Instead it is used by oligo to determine which probes should be collected into a given probeset when running RMA.

If you are summarizing at the core (transcript) level, then you need the huex10sttranscriptcluster.db package to do the annotations. However, see Rebuilding annotation packages for Affymetrix Gene/Exon ST Arrays.
 

Hi Elena,

When the current Bioconductor version was released, the annotation version available from Affymetrix was NA33.1... This is the version used to create the package you're using.

Affymetrix released the NA35 annotations after our Bioc release.

You can download the (source) package from the link below:

https://www.dropbox.com/s/2mamf738h0svmyc/pd.huex.1.0.st.v2_3.13.0.tar.gz?dl=0

benilton

Dear James and Benilton,

thank you a lot for your answers, and thank you, Benilton, for the link to the package.

I've made one of  my own (it was the first time I tried it), and, although I've got a package, and was able to install it and work with it , some of the information present in the pd.huex.1.0.st.v2_3.12.0 was not anymore in my package and I was struggling to know why it was so.

Thank you again!!

Best regards,

Elena

Hi Elena,

So we have been digging into this since you brought it to our attention, and it seems that the set of Affy based annotations that were recommended by Affy as the most recent at the time when the last batch of annotation packages were made (IOW all of the packages based on Affys na34 release) have since been declared to be 'sub-par' by Affy and replaced with newer na35 annotations.  We all care very strongly about making sure that these are the best that they can be, and so we are going to be replacing them over the next week with fresh packages based on na35.  I already have updates for Jims packages (they are being checked as I write this), and I am rebuilding the ones that we normally make here).  Beniltons will also be rebuilt, but his take a couple more days to finish, so those should be online later next week.

As they come out, you will be able to update by just using biocLite()

Thanks for bringing this to our attention!

 Marc