Genetical Research



Extending the loop design for two-channel microarray experiments


NAOMI S. ALTMAN a1c1 and JUN HUA a1
a1 Department of Statistics, Thomas Building, The Pennsylvania State University, University Park, PA 16801-2111, USA

Article author query
altman ns   [PubMed][Google Scholar] 
hua j   [PubMed][Google Scholar] 

Abstract

The loop design of Kerr and Churchill is a clever application of incomplete blocks of size 2 to two-channel microarray experiments. In this paper, we extend the loop design to include more replicates, biological and technical replication, multi-factor experiments, and blocking. Loop and extended loop designs are shown to be more efficient than the reference design for any given number of arrays. We also show that adding new treatments to a loop design requires the same number of additional arrays as adding treatments to a reference design, with a greater gain in power. Given the flexibility of extended loop designs and their power, we propose that these should be the designs of choice for most experiments using two-channel microarrays.

(Received July 25 2006)
(Revised November 16 2006)


Correspondence:
c1 Telephone: +1 (814) 965 3791. Fax: +1 (814) 863 7114. e-mail: naomi@stat.psu.edu


Metrics