Correspondence analysis of genes and tissue types and finding genetic links from microarray data.

摘要: In this paper, we propose and use two novel procedures for the analysis of microarray gene expression data. The first is correspondence which visualizes relationship between genes tissues as 2 dimensional graphs, oriented so that distances are preserved, primarily distinguish certain types tissue spatially close to those tissues. For inference genetic links, partial correlations rather than key issue. A correlation i j after effect surrounding has been subtracted out their pairwise correlation. This leads area graphical modeling. limitation modeling approach matrix profiles degenerate whenever number be analyzed exceeds distinct measurements. can cause considerable problems, calculation typically uses inverse matrix. To avoid limitation, practical multiple regression with variable selection measure net, screened, pairs genes. Possible biases arising from a subset genome examined in worked examples. It seems both these approaches more natural ways analyzing data currently popular way clustering.