[BioC] Microarray slide design and data analysis

[William Kenworthy]

Hi, 
we have just bought in a few microarray slides and they have each gene
characterised by only a single spot.  This is not really desirable as I
would like to see (ideally) a triplicate or at least a duplicate of
spots distributed for each gene to allow some kind of quality check
(variability) as well as use some of the more advanced analysis tools on
them.

Can anyone point me to any published papers or working papers on slide
design that cover these points:
* single/duplicate/triplicate/many spots
* advantages of replicate slides and dye swaps 
* comparison of side-by-side and distributed spot patterns (i.e., are
distributed a la affy spot patterns any better than localised spots:
spatial effects, artifacts and variability across the slide are some of
the concerns I can think of)
* Controls (scorecard and the like, housekeeping gene's, lighthouse
spots, blanks, buffer spots etc) - this is a huge topic and there are a
few papers out there on this, but if someone wants to comment ...

Searches turn up a lot of empirical data and a few gems, but I would
like to find more references to support an "ideal slide design for data
analysis", to counter the idea that slide real-estate is expensive, so
we must therefore crowd as much on a slide as possible so we can cover
the largest number of gene's possible!

Your thoughts please ...

BillK

-- 
William Kenworthy
BioInformatics Officer
  Centre for BioInformatics and Biological Computing
  WA State Lotteries Microarray Facility
W.Kenworthy at murdoch.edu.au
+61 (0)8 9360 2790