Biclustering with heterogeneous variance

摘要: In cancer research, as in all of medicine, it is important to classify patients into etiologically and therapeutically relevant subtypes improve diagnosis treatment. One way do this use clustering methods find subgroups homogeneous individuals based on genetic profiles together with heuristic clinical analysis. A notable drawback existing that they ignore the possibility variance gene expression profile measurements can be heterogeneous across subgroups, not consider heterogeneity lead inaccurate subgroup prediction. Research has shown hypervariability a common feature among subtypes. paper, we present statistical approach capture both mean structure data. We demonstrate strength our method synthetic data two sets. particular, confirms methylation level patients, detects clearer patterns lung