Construction of Efficient Block Designs for Microarray Experiments andMixed Effects

Microarray experiments form an important tool in gene expression analysis. If arrays are considered as blocks and varieties as treatments (e.g. type of tissues, drug treatments), then a classical incomplete block design can be usefull for microarray experiments. Construction of optimal designs under some of the optimality criteria for these experiments is very important. But, optimal block designs with theoretical methods are difficult to obtain. Therefore, finding optimal block designs using computeraided searches has received a lot of attention. In this paper, the treatment effects and the block effects are taken to be fixed and random, respectively. Then an algorithm is proposed to obtain the optimal or near-optimal designs in a short time even for large numbers of blocks or treatments or both. Also, the algorithm is developed for k-colour.