Hello,everyone! We know that beadarray can handle bead-level data, but it seems that there is no way for lumi to do this.

lumi could read 'summarized intensities' in, just as follows:

#summarized expression profile for two samples
ProbeID    AVG_SIGNAL-1    DETECTION-1    AVG_SIGNAL-2    DETECTION-2
ILMN_1762337    56.86815    0.9538867    67.17416    0.3241107
ILMN_1736007    71.69266    0.3254282    73.8208    0.1172596<font face="sans-serif, Arial, Verdana, Trebuchet MS">
</font>

but no way to read in this directly:

#a bead-level text file for one sample, contains 'array_address_id intensity locX locY'
Code    Grn    GrnX    GrnY
10008    2530    697.5355    3672.499
10008    2881    1112.538    6927.974

I had tried to read in this kind of files by lumiR or lumiR.batch by extracting the 'array_address_id' and 'intensity'-'Code' and 'Grn' above, it seems to be worked! But I couldn't prove that it is a right way.

Any idea for these?

Hello,

I am also working with beadarray illumina data and I pre-process them using lumi. I do not read directly the beadarray files however; I first upload the raw data to Illumina Genome Studio (IGS) and then I export the sample and control probe information to csv files...which I then read using lumi. Do you have IGS and can you use it?

Best,

Eleni

Hi,

The beadarray and lumi packages are aimed at working with different levels of data.  As you've already found out, beadarray can work with bead-level data (where we have intensities for every bead on the array), but lumi is intended only to work with summarised data (a single average intensity for each bead-type).

Personally I think there are advantages to starting work with the bead-level data if you have it, where the ability to know the location of beads allows you to exclude or correct sections of an array where something has gone wrong during the imaging. You can also perform more robust outlier calling than the default Illumina methods.  However ultimately you then do summarize to a single value per bead-type. The summarize() function in beadarray will get you to this stage.

From here you could export the values to a text file and then read them into lumi - there may be some manual work required here.  Typically I would go straight to limma to perform a differential expression analysis.