Hi all,

I've processed some differential analysis by limma on Illumina Gene Expression Array, and I'm working to intersect the results.

I am of the opinion that the correct choice to intersect is by Probe Id (unique field) and I usually did so.
Now I was wondering again if you want to intersect by Gene Symbol what might be the best approach, in the sense that as we know more probes refer to the same Gene.

So possible scenarios:

1)I can have between lists two probes with different Probe Id that refer to the same Gene

2)or more probes in a list, more probes in the other list that refer to the same Gene and a possible criterion of choice for a match is to select the one with the best adj.P.Val or B-statistic (lods) in both lists.

I always stand in my opinion to intersect by unique Probe Id field, but sometimes may come a few matches and it may be useful extend the output proceeding by Gene Symbol.

Thanks in advance,
Giovanni