Hi, More information about the probe-names problems I'm having with probe names becoming X12324.at instead of 1234_at Firstly, this is the version of R I'm using: >R.version _ platform i686-pc-linux-gnu arch i686 os linux-gnu system i686, linux-gnu status Beta major 1 minor 7.1 year 2003 month 06 day 08 language R I've managed to reproduce the wierdness with the following minimal script: I start off with two dataframes: class(first); class(second); > class(first); [1] "data.frame" > class(second); [1] "data.frame" containing some numbers calculated for each probeset (these are both generated using identical functions, just comparing different groups of chips): first[1:2,] second[1:2,] > first[1:2,] X1 X2 X3 X4 1007_s_at 0 0.1218871 0.2664602 1.202853 1053_at 0 0.2757304 0.4460427 1.362298 > second[1:2,] X1 X2 X3 X4 1007_s_at 0 0.6255407 -0.08225843 -1.058674 1053_at 0 0.6037322 -0.19470686 -1.144492 I then find entries where X2<0.05 in each data.frame: > first[first$X2<0.5,][1:2,] X1 X2 X3 X4 1007_s_at 0 0.1218871 0.2664602 1.202853 1053_at 0 0.2757304 0.4460427 1.362298 > second[second$X2<0.5,][1:2,] X1 X2 X3 X4 X121.at 0 0.09278053 -0.19887301 -1.147801 X1255.g.at 0 0.41586962 0.09071178 1.064895 Weird! Now, I play a bit: > second[1:2,][1:2,] X1 X2 X3 X4 1007_s_at 0 0.6255407 -0.08225843 -1.058674 1053_at 0 0.6037322 -0.19470686 -1.144492 > second[1:2,][1:3,] X1 X2 X3 X4 X1007.s.at 0 0.6255407 -0.08225843 -1.058674 X1053.at 0 0.6037322 -0.19470686 -1.144492 NA NA NA NA NA > Does this give a clue as to what's going on? Cheers, Crispin -------------------------------------------------------- This email is confidential and intended solely for the use of th... {{dropped}}

> > second[1:2,][1:2,] > X1 X2 X3 X4 > 1007_s_at 0 0.6255407 -0.08225843 -1.058674 > 1053_at 0 0.6037322 -0.19470686 -1.144492 > > second[1:2,][1:3,] > X1 X2 X3 X4 > X1007.s.at 0 0.6255407 -0.08225843 -1.058674 > X1053.at 0 0.6037322 -0.19470686 -1.144492 > NA NA NA NA NA > > for this part, you can see what is happening by doing debug(get("[.data.frame")) and running your second subscripting operation. make.names is invoked when any row name is NA. it may be worth posting this to r-help. it is not really a bioconductor issue.