9^th World Biomarkers Congress

Recently, the researches to fi nd proteins as biomarkers of various diseases have become widespread. Especially these studies are popular in оncology. However, until now proteins specifi c for cancer cells are not detected. In spite of it, to diagnose cancer diseases biomarkers which are specifi c for a neoplastic process generally are widely used (such as CA-125, CA 19-9, CA 15-3, PSA, α-fetoprotein etc.). However, they can’t discharge the main task: to diagnose the initial stages of disease, because their opportunity is limited by low sensitivity of existing laboratory technologies. Th e situation is much better in the case of using RNA’s as biomarkers, because PCR takes off such restrictions. In the case of metabolomic diagnostics there are the same concentration limits as in proteomics. Th erefore, it was decided not to use separate low molecular weight compounds as biomarkers but the whole metabolomic profile of the sick person as a biomarker itself of a disease. With this purpose, at the fi rst stage that profi le was clusterized using the method of principle component analysis to detect clusters of metabolites which are diff erent from the clusters of healthy person. At the second stage, the specifi city and sensitivity of the proposed technology was evaluated using the SVM and ROC analyses, for diagnostics of lung, prostate and other cancers.