Hi Fabio, As the Illumina and its BeadStudio software keep changing the Identifiers and output format, it makes difficult to combine the datasets in different batches. That is why we recommend to use nuID as the probe identifiers. There are two ways to convert Illumina Ids to nuIDs: 1. from ProbeID to nuID; 2. from TargetID to nuID. We recommend using the first conversion because ProbeID is unique for each probe, also because Illumina changed TargetIDs as shown in your question. In order to using ProbeIDs, you need to input data from the ³Sample Probe Profile.txt² or ³Group Probe Profile.txt² outputted by BeadStudio. The ³Sample Gene Profile.txt² is identified by TargetIDs. We do not recommend using this file because it combines the probes corresponding to the same gene (the probes for the same genes may have quite different expression profiles). So I suggest you input the data again using ³Sample Probe Profile.txt² and convert them to nuID, then you can combine them if they have the same nuIDs. I just updated the ³combine² function to make sure the probes have the same order. I will send them to you in a separate email or you can update the lumi package (lumi 1.5.17) in the next few days. Thanks for reporting the problem. Pan On 2/10/08 6:31 AM, "fabio manzo" wrote: > Dear Pan Du, > I disturb you again for another problem that I'm not able to solve alone. I > got two Illumina dataset with two different targetID, one the > GI_10047089-S > GI_10047091-S > GI_10047093-S > GI_10047099-S > GI_10047103-S > GI_10047105-S > GI_10047121-S > GI_10047123-S > GI_10047133-A > GI_10047133-I > GI_10092578-S > and the other > ILMN_10000 > ILMN_10001 > ILMN_10002 > ILMN_10004 > ILMN_10005 > ILMN_10006 > ILMN_10009 > ILMN_1001 > ILMN_10010 > ILMN_10011 > I tried to import both with lumiR and asking for making nuID with library > lumiHumanV1 and it's working lovely with the first one but less with the > second one. Do I need to apply the function probeID2nuID(lumi)? > My second problem is how to fuse these two dataset: > I tried to "combine" them, but R told me that's not possible because they have > different row.names and different dimension, either like dataframe either like > matrix. I was thinking if it' s possible implement a redundant function that > can compare the two row.names, that just those are equal and then copy > them in a new dataframe with their rows. Finally combine them. A sort of: > A =nuID gene first dataset > B= nuID gene second dataset > if A=B, > read () > "copy" (in a new dataframe) > then make a unique dataframe like x= data.frame(common nuID, first dataset, > second dataset), because now they should have the same names, same order, same > dimensions, at least like number of rows. > I'm trying to study either the R languange either the C language, but even in > this case I'm really a beginner and it could take like one year just to be > able manage the languages and ...I'm not sure to succeed. I would like to ask > you if a function like that could be put on the lumiR.batch, that also gave me > the same error or if I can avoid this problem in other ways. > I'm sorry if I wrote you a sort of question's book and for all the time and > the patience lost. > Thanking you for all your help, > Best > fabio > > --------------------------------------------------- Pan Du, PhD Research Assistant Professor Robert H. Lurie Comprehensive Cancer Center Northwestern University 676 ST Clair St., #1200 Chicago, IL 60611 Office (312)695-4781 dupan@northwestern.edu --------------------------------------------------- [[alternative HTML version deleted]]