Hi all. I am using HG-U133 PLUS 2  platform. I am having one problem of one  probe set ID mapping to several genes? How can one resolve this problem? is there a bioconductor software that deals with this problem? Thank you.

There is no easy resolution to the 'problem', as there are multiple reasons that a probeset may be annotated to multiple genes:

• The genes may be so closely related that Affy couldn't find a set of 25-mers that would reliably distinguish between the two
• There may be a miRNA within the binding region of the probeset, so it isn't possible to say if binding of target is due to the translated gene, or the miRNA
• Some subset of the probes in a probeset may bind to multiple highly repetitive regions of the genome
• What looks like multiple genes is simply duplication in one or more databases that has not been resolved yet (e.g., a given gene may have multiple Entrez Gene IDs, Ensembl IDs, HUGO symbols, etc, which will eventually get sorted out)

Since there are multiple reasons for duplication, it isn't possible to 'fix' the issue with a simple rule. Well, that's not exactly true. You can use mapIds(), which has a 'multiVals' argument that you can use to deal with duplicates, and you can decide upon a simple rule like 'just take the first one' or 'randomly select one' or 'just give me a NA for those'. But that only works for one mapping at a time. The alternative is to use select() and then split() to convert to a list, and then select individual rows, depending on how you want to choose.