Hi ! =20 I am trying to reanalyze the data from de Gregorio et al. (http://www.fruit= fly.org/expression/immunity/) using the affy package, wich I have already s= uccessfully used with some other datasets. However I am encountering a stra= nge problem here. ReadAffy seems to have a problem with 4 of the 34 chips o= f the set. =20 Here is the error message for the four chips : =20 > test <- ReadAffy(widget =3D TRUE) Warning messages:=20 1: could not find CDF name, setting it to 'unknown' in: whatcdf(file, compr= ess =3D compress)=20 2: could not find CDF name, setting it to 'unknown' in: whatcdf(file, compr= ess =3D compress)=20 3: could not find CDF name, setting it to 'unknown' in: whatcdf(file, compr= ess =3D compress)=20 4: could not find CDF name, setting it to 'unknown' in: whatcdf(file, compr= ess =3D compress)=20 When I manually open the files I don't see any difference between the 'good= ' and the 'bad' files : =20 a good file's headers : [CEL] Version=3D3 [HEADER] Cols=3D640 Rows=3D640 TotalX=3D640 TotalY=3D640 OffsetX=3D0 OffsetY=3D0 GridCornerUL=3D226 233 GridCornerUR=3D4488 233 GridCornerLR=3D4488 4498 GridCornerLL=3D227 4499 Axis-invertX=3D0 AxisInvertY=3D0 swapXY=3D0 DatHeader=3D[39..46276] bacterial infection t0 array1:CLS=3D4733 RWS=3D473= 3 XIN=3D3 YIN=3D3 VE=3D17 2.0 02/20/ 1 12:23:18 =14 GridVerify= =3DNone =14 DrosGenome1.1sq =14 =14 =14 =14 =14 =14 =14 =14 =14 6 Algorithm=3DPercentile AlgorithmParameters=3DPercentile:75;CellMargin:2;OutlierHigh:1.500;Out lierL= ow:1.004 [INTENSITY] NumberCells=3D409600 =20 The same for a 'bad' file : [CEL] Version=3D3 [HEADER] Cols=3D640 Rows=3D640 TotalX=3D640 TotalY=3D640 OffsetX=3D0 OffsetY=3D0 GridCornerUL=3D225 234 GridCornerUR=3D4488 240 GridCornerLR=3D4482 4505 GridCornerLL=3D220 4500 Axis-invertX=3D0 AxisInvertY=3D0 swapXY=3D0 DatHeader=3D[49..46275] bacterial infection t0 array2 scan2:CLS=3D4733 RWS= =3D4733 XIN=3D3 YIN=3D3 VE=3D17 2.0 02/20/ 1 16:05:31 =14 GridV= erify=3DNone =14 DrosGenome1.1sq =14 =14 =14 =14 =14 =14 =14 =14 = =14 6 Algorithm=3DPercentile AlgorithmParameters=3DPercentile:75;CellMargin:2;OutlierHigh:1.500;Out lierL= ow:1.004 [INTENSITY] NumberCells=3D409600 =20 Quite logically, justRMA is just as confused with those files. Youcan downl= oad some the CEL files at the following address : http://134.59.90.101/chip= s/chips.htm =20 Any advice, question, explanation would be more than welcome ! =20 -- Fabrice Girardot Equipe Biologie Virtuelle UMR 6543 Nice, France --------------------------------- [[alternative HTML version deleted]]

Hi again, =20 My problem was effectivelly a too long line in the headers of the .cel file= s... My install of R and Bioconductor is up to date, so it should have the = modif that you mention (but I didn't check that). What I did was simply to = reduce the size of the descriptive field in the headers from : DatHeader=3D[32..46261] bacterial infection t0 array2 scan2[...]=14 DrosGe= nome1.1sq =14 =14 =14 =14 =14 =14 =14 =14 =14 6 to : DatHeader=3D[49..46275] b inf t0 array2 scan2[...]=14 DrosGenome1.1sq =14 = =14 =14 =14 =14 =14 =14 =14 =14 6 and now it works nicely ! =20 Simon, thanks a lot ! =20 -- Fabrice Girardot Equipe Biologie Virtuelle UMR 6543 Nice, France --------------------------------- [[alternative HTML version deleted]]