| 摘要 |
PROBLEM TO BE SOLVED: To structure a genealogical tree upon the basis of frequency data as to many modifications detected with a sample of cancer according to the present invention.SOLUTION: Analysis is taken upon the basis of data (hereinafter referred to as modification group frequency data) in which samples with modifications close in frequency among a plurality of samples in which a plurality of clones having different genomes are mixed are grouped, and frequencies of modifications of the respective groups are enumerated. While it is considered that clones corresponding to modification groups in which all samples have the same large/small relation in modification group frequency have parent-child relations, a graph structure is formed which includes clones as vertices and parent-child relations as branches, parent-child relations of clones contradicting modification group frequency data are erased from this graph, and further clones as parents of respective clones are selected while taking correlation coefficients of modification group frequencies of a plurality of samples into consideration.SELECTED DRAWING: Figure 1 |
