. . . . . . . "[Multinomial logistic regression models based on the CGI methylation profiles from normal mucosa correctly identified 78.9% of cancer patients and 87.9% of non-cancer (neoplasia-free+polyp) patients (P=4.93 x 10(-7)) using APC, HPP1, p16, SFRP4, WIF1 and ESR1 methylation as the most informative variables.]. Sentence from MEDLINE/PubMed, a database of the U.S. National Library of Medicine."@en . . . . . "2017-02-19"^^ . . "Gene-disease associations inferred from text-mining the literature."@en . "DisGeNET evidence - LITERATURE"@en . "2017-10-17T13:10:29+02:00"^^ . . . . . . . . . . . "v5.0.0.0" . "v5.0.0" .