I am analyzing two color microarrays with limma but I am getting blank and empty spots as differentially expressed genes in topTable. I have made spot files but can't get get rid of them. Any help will be appreciated.

> I am analyzing two color microarrays with limma but I am getting blank > and empty spots as differentially expressed genes in topTable. I have > made spot files but can't get get rid of them. You can use logical subsetting of your MA or RG object. E.g. you could use something like RG = RG$genes$Status = controlStatus(spottypes, RG) RGnoblanks = RG[!(RG$gene$Status %in% c('blank','empty')), ] The only catch with this is that somehow the color attribute seems to get lost in the process, resulting in a surprise the next time you use plotMA(). But you can fix this by re-doing the controlStatus step. (BTW: does anybody here know why the color attribute is dropped?) The other side of the story is, that blank spots should not appear to be differntially expressen in the first place... cu Philipp -- Dr. Philipp Pagel Tel. +49-8161-71 2131 Lehrstuhl f?r Genomorientierte Bioinformatik Fax. +49-8161-71 2186 Technische Universit?t M?nchen Wissenschaftszentrum Weihenstephan 85350 Freising, Germany and Institut f?r Bioinformatik / MIPS GSF - Forschungszentrum f?r Umwelt und Gesundheit Ingolst?dter Landstrasse 1 85764 Neuherberg, Germany http://mips.gsf.de/staff/pagel

Quoting Narendra Kaushik <kaushiknk at="" cardiff.ac.uk="">: > I am analyzing two color microarrays with limma but I am getting > blank and empty spots as > differentially expressed genes in topTable. I have made spot files > but can't get get rid of them. > > Any help will be appreciated. There are several reasons why you may get blank/empty spots as differentially expressed. First of all, the blank/empty spot may not be empty after all. My CRUK22K (cDNA arrays) were like that. When I asked the makers about it, as I was worried I had the wrong gene list, they told me that after they started printing the arrays they found some cDNAs were not reliable (might have been misidentified, or simply lost track of them), and they renamed them on teh gene list as "blank", rather than something else... I do hope this practice is not common, but it can happen! However, the most common reason is probably that you have a real blank spot, with near background intensities (very very low). The signal ratios for low intensity spots can be quite high, depending on the background correction method used, and may give rise to misleading results like these. Do your apparently DE blank/empty spots also have a low A value? My guess is yes. You can simply filter your topTable, eliminating any blank/empty spots from the list. I personally don't use SpotFiles. When I want to highlight spots, I make a vector of the unique IDs identifying them, and use that instead. I find it more simple/flexible that way. Jose -- Dr. Jose I. de las Heras Email: J.delasHeras at ed.ac.uk The Wellcome Trust Centre for Cell Biology Phone: +44 (0)131 6513374 Institute for Cell & Molecular Biology Fax: +44 (0)131 6507360 Swann Building, Mayfield Road University of Edinburgh Edinburgh EH9 3JR UK -- The University of Edinburgh is a charitable body, registered in Scotland, with registration number SC005336.