[R] gcrma and chip without mm values

[Karin Lagesen]

I am working with a custom chip which does not have mm probes, 
but which does have negative control probes. I would like to 
see the results of gcrma adjustment of the probe values.

I have thus chosen all of the indices of the negative controls 
and am trying to use them as described in the gcrma package:

The indices are here:


> randomprobeindexes[1:5,]
       x   y
[1,] 114 117
[2,] 116 117
[3,] 118 117
[4,] 120 117
[5,] 122 117
>

I now use them to do gcrma:

> bg.adjust.gcrma(newdata, affinity.source="local", type="affinities", NCprobe=randomprobeindexes)
Adjusting for optical effect.Error in exprs(abatch)[Index, i] <- exprs(abatch)[Index, i] - min(exprs(abatch)[Index,  : 
        NAs are not allowed in subscripted assignments
> 

This is due to this:

> mm(newdata)[200:205,]
     dmso5.CEL mnng5.CEL mock5.CEL uv5.CEL
[1,]        NA        NA        NA      NA
[2,]        NA        NA        NA      NA
[3,]        NA        NA        NA      NA
[4,]        NA        NA        NA      NA
[5,]        NA        NA        NA      NA
[6,]        NA        NA        NA      NA
> 

And:

> mmindex(newdata)[200:202]
$`AFFX-yel006_5_copy5_at`
 [1] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

$`AFFX-yel006_5_copy6_at`
 [1] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

$`AFFX-yel006_M_copy1_at`
 [1] NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA NA

> 

We do have some mm values on the chip, since there is always one
control set present which does have mm probes with it. This one
is required for affymetrix own purposes. Besides this there are
no mm values. which means that most of the mmindex list just
states na, and as shown above, so does the mm list too.

Now, my initial strategy was to try and replace all mm values 
with zero. gcrma does not use mm values when negative control
indices are given, so this should, afaik, be ok. However, I
am not able to do this, and I think this is due to me not 
having any mm indices to use for the replacement.

Does any of you have any tips on how to overcome this?

TIA,

Karin
-- 
Karin Lagesen, PhD student
karin.lagesen at medisin.uio.no
http://folk.uio.no/karinlag