Using KNIME for Gene Expression analysis: how to read CEL files in KNIME ?

KNIME Analytics Platform
1

I am participating at the SBVIMPROVER competition on diagnostic signature detection (www.sbvimprover.com), that is basically  a data mining competition in biomedical field. It uses published gene expression datasets for training and provides "unlabeled" for benchmarking.

I therefore thought to use KNIME for that competition, however there is a problem due to the format of the data, so called CEL files from Affymetrix.

Although one can read CEL files using R packages,  one would have to do that outside KNIME and that is not ideal (particularly when dealing with many input files (.

Is there anyone that has already developed or is developing nodes to read CEL files ?
If not, then I could involve some local resources to do that, and contribute the node to the community.

Looking forward your advice / reply.

2

If this functionality already exist in R, I would suggest making use of it and use the R integration in KNIME available on our in-build update site.

I would recommend the following setup in KNIME: use the List Files node to read all files from a directory (with a certain extension) and provide them to the Chunk Loop Start node (chunk-size: 1) and provide each of the files to the R Snippet which runs the R script to read this particular format. Note, the input data (simply a file name) is available within a data.frame called R and the final data need to be return inside an variable R again. To close the loop, use the Loop End node which collects all the different files (data chunks).