Biclustering of expression data with evolutionary computation

摘要: Microarray techniques are leading to the development of sophisticated algorithms capable extracting novel and useful knowledge from a biomedical point view. In this work, we address biclustering gene expression data with evolutionary computation. Our approach is based on algorithms, which have been proven excellent performance complex problems, searches for biclusters following sequential covering strategy. The goal find maximum size mean squared residue lower than given /spl delta/. addition, pay special attention fact looking high-quality large variation, i.e., relatively high row variance, low level overlapping among biclusters. quality found by our discussed results compared those reported Cheng Church, Yang et al. general, approach, named SEBI, shows an at finding patterns in data.