Multi-state genes are ones with multi-modal expression patterns (i.e. expression of discretized states) or multi-modal patterns. For each probeset, CNS regions were grouped into the several "states" according to their expression values with variational Bayesian inference to fit to a gaussian mixture model. Multi-state genes, which have two or more states, were shown in the list. one-state genes, which have only one state, were not shown in this view.
States of multi-state genes
According to the order of mean values of every states, "high"-, "low"-, "up"-, and "down"-states are defined. "high"-state has regions in the state with the maximum mean expression value, "low"-state has regions in the state with the minimum mean expression value, "up"-state has regions both in "high"- and middle-states, and "down"-state has regions both in "low"- and middle-states.
Go to the documents page