Dear Adai, Thanks a lot for your answer. You actually provide a lot of interesting suggestions, and I'll ask about some of these possible options. Yes, I looked at the nature Methods and the PNAS paper, but some of the platforms I was interested in were not included. The people I've talked to, they all seemed happy with the platform they were using, and that is why I was looking for possible comparisons. Best, R. On Tuesday 17 May 2005 21:02, Adaikalavan Ramasamy wrote: > Sorry, I have no experience with any of these arrays. > > If you can spend a bit of money, then it might be worth investigating > their reproducibility (and perhaps the investigation might end up as an > article too). > > I think you might be able to send your samples to these different > companies and ask them to do the hybridisation for you explaining that > you are interested in long term commitment if the hybridisation works > out. They may offer to do so at a special price. > > Then unknown to them send some sample along with some repeat aliquots > (e.g. 5 biological samples x 3 aliquots = 15 "samples"). This way you > can assess the biological and technical variability when it is done > under the best procedures. > > If you have more money, then give the same samples to someone who is > relatively new with microarray (e.g. PhD student?). Then calculate the > variability and compare with above. This would tell you how sensitive > each platform is to bad handling. > > A modification of this is to spike in some genes of known concentration > or supplying the same sample to many different labs. > > The following news articles covers 3 recent studies that compared the > effect of different platforms and labs. Try scanning them to see if the > platforms you mentioned are covered by these studies > http://focus.hms.harvard.edu/2005/May6_2005/comp_bio.shtml > > If you do not have the money, then do a literature search and talk to > many people who have done the array. > > Regards, Adai > > On Tue, 2005-05-17 at 18:54 +0200, Ramon Diaz-Uriarte wrote: > > Dear All, > > > > We are trying to decide on whether to use Agilent two-color arrays, > > Codelink arrays or Applied Biosystems arrays (the last two are one-color) > > for a new study. An advantage of the one-color platforms is that > > design-related issues (loop, reference, all pairwise, etc) are no longer > > an issue. This is very relevant in our case because we have 5 groups of > > very unequal sample sizes and we?ll need to correct for the effect of > > several covariates, so finding an optimal design seems difficult. > > > > I've seen several recent posts in the lists concerning Codelink arrays. I > > wonder if anybody has any comparison among these platforms (specially if > > forced to use a common reference design for the two-color arrays), or any > > experience with the Applied Biosystems arrays. > > > > Best, > > > > R. -- Ram?n D?az-Uriarte Bioinformatics Unit Centro Nacional de Investigaciones Oncol?gicas (CNIO) (Spanish National Cancer Center) Melchor Fern?ndez Almagro, 3 28029 Madrid (Spain) Fax: +-34-91-224-6972 Phone: +-34-91-224-6900 http://ligarto.org/rdiaz PGP KeyID: 0xE89B3462 (http://ligarto.org/rdiaz/0xE89B3462.asc) **NOTA DE CONFIDENCIALIDAD** Este correo electr?nico, y en ...{{dropped}}