[BioC] affy plus2 array - unclear probeset signal

andrea.grilli at ior.it andrea.grilli at ior.it
Mon May 20 14:38:51 CEST 2013


Dear BioC members,
we profiled a cell line (parental) in a time-series experiment,  
comparing it with the same cell line transfected by a gene. We  
verified the transfection by real-time PCR and it was ok.
When we moved to gene profiling (HG-U133a plus2 array), I discovered  
that the 2 probesets matching the transfected gene have: 1st probeset,  
same high expression in parental and transfected cells (about 12.5 of  
log2 absolute values after RMA normalization), 2nd probeset, strongly  
higher expression in parental cell line, so the opposite than attended.

Premising that:
- we hadn't inversion of the samples (real-time PCR was done on same RNA).
- probesets don't cross-hybridize to other genes (according to NetAffx  
tool and also blast analysis).
- the 2 probesets match different part of the transcript: 1st probeset  
only the 3'UTR of the transcript, instead 2nd probeset different exons.
- No different isoforms of the gene are matched by the 2 probesets.
- I discovered a publication* talking about the probesets reannotation  
on the same array: they indicated that one probeset (8/10 probes to be  
more precise) casually matches the SV40 poly a region, which is in the  
vector we used for transfection. The signal we have from this probeset  
is exactly what we expected, with higher and increasing expression  
during time in transfected cells (so supporting the idea of no samples  
inversion).

Is there anything else that I can check to unravel the problem? Should  
I consider one of the two probes more reliable, or discard the 2nd  
"wrong" probe beforehand? Any suggestion on how to explain the problem?

Thanks in advance,
Andrea



* A sequence-based identification of the genes detected by probesets  
on the Affymetrix U133 plus 2.0 array. Harbig, NAR, 2005.


Dr. Andrea Grilli
andrea.grilli at ior.it
phone 051/636.6756-6937

Laboratory of Experimental Oncology,
Development of  Biomolecular Therapies unit,
Rizzoli Orthopaedic Institute
Codivilla Putti Research Center
via di Barbiano 1/10
40136 - Bologna - Italy



More information about the Bioconductor mailing list