Combining Results of Microarray Experiments: A Rank Aggregation Approach

摘要: As technology for microarray analysis becomes widespread, it is becoming increasingly important to be able compare and combine the results of experiments that explore same scientific question. In this article, we present a rank-aggregation approach combining from several studies. The motivation twofold; first, final studies are typically expressed as lists genes, rank-ordered by measure strength evidence they functionally involved in disease process, second, using information on metric means do not have concern ourselves with data actual expression levels, which may comparable across experiments. Our draws methods top-k computer science literature meta-search. meta-search problem shares features experiments, including fact there few many elements common all lists. We implement two algorithms, use Markov chain framework convert pairwise preferences between list into stationary distribution represents an aggregate ranking (Dwork et al, 2001). behavior algorithms hypothetical examples simulated dataset their performance algorithm based order-statistics model Thurstone (Thurstone, 1927). apply three five prostate cancer.