초록 close

Receiver operating characteristic (ROC) approach can be employed torank candidate genes from a microarray experiment, in particular, for thebiomarker development with the purpose of population screening of a cancer.In the cancer microarray experiment based onn patients the researcher oftenwants to compare the tumor tissue with the normal tissue within the sameindividual using a common reference RNA. Ideally, this experiment producesn pairs of microarray data. However, it is often the case that there aremissing values either in the normal or tumor tissue data. Practically, wehave n1 pairs of complete observations,n2 \normal only" and n3 \tumoronly" data for the microarray. We refer to this data set as a mixed dataset. We develop a ROC approach on the mixed data set to rank candidategenes for the biomarker development for the colorectal cancer screening. Itturns out that the correlation between two ranks in terms of ROC andtstatistics based on the top 50 genes of ROC rank is less than 0.6. Thisresult indicates that employing a right approach of ranking candidate genesfor the biomarker development is important for the allocation of resources.