Mining Disease Fingerprints From Within Genetic Pathways

摘要: Mining biological networks can be an effective means to uncover system level knowledge out of micro associations, such as encapsulated in genetic pathways. Analysis human disease pathways lead the identification major mechanisms that may underlie disorders at abstract functional level. The focus this study was develop approach for structural pattern analysis and classification diseases. A probabilistic model developed capture characteristic components (‘fingerprints’) functionally annotated probability estimation procedure searched fingerprints each pathway while improving estimates parameters. evaluated on data from Kyoto Encyclopedia Genes Genomes (consisting 56 across seven categories). Based achieved average accuracy up ~77%, findings suggest these used discovery