[BioC] Affymetrix - index of probes

[Laurent Gautier]

Just a (side) comment.

The probe set with 69 probes (see James' very demonstrative example), is 
an odd one: all its probes are designed to match ALU repeats. The probe 
set does not monitor the expression level of a particular gene (and 
computing the expression value for it does not necessarily make a lot of 
sense).


L.



James W. MacDonald wrote:
> Benjamin Haibe-Kains wrote:
> 
>> Hi,
>>
>> I want to compare affy hgu133a and hgu133plus2 chips. When I retrieve 
>> the index of different probes on these two chips, the function 
>> indexProbes (see affy package) returns a list of 16 probes for the pm 
>> and 16 probes for the mm for each probe set.
> 
> 
> Not exactly true. There are some probesets with greater or less than 11 
> PM probes, but in general there are 11 per probeset.
> 
>  > cdfName(dat)
> [1] "HG-U133A"
>  > table(unlist(lapply(indexProbes(dat, "pm"), length)))
> 
>     8    10    11    13    14    15    16    20    69
>     1     1 21748     4     4     2   482    40     1
> 
>  > cdfName(dat2)
> [1] "HG-U133_Plus_2"
>  > table(unlist(lapply(indexProbes(dat2, "pm"), length)))
> 
>     8     9    10    11    13    14    15    16    20    69
>     5     1     6 54130     4     4     2   482    40     1
> 
> HTH,
> 
> Jim
> 
> 
> 
>>
>> However, we can see in the Affymetrix doument 
>> http://www.affymetrix.com/support/technical/datasheets/human_datasheet.pdf 
>>
>> that there are only 11 probes per probe set ! I don't understand the 
>> origin of the 5 exceeded probes ...
>>
>> Can you help me ?
>>
>> thanks
>>
>> NB: this email is maybe duplicated ...
>>
> 
>