| Rencently, biomarkers become a hot topic in the study on disease diagnosis and drug effectiveness evaluation, especially in some thorny issues in clinic, like the early diagnosis of diabetic nephropathy, a renal complication of diabetes. In this dissertation, an integrated biomarker system was developed using clinical information and multilevel laboratory data. Then some screening methods, like artifical intelligence techniques, significance test and ROC analysis were applied to obtain the optimal biomarker for clinical challenges.1. Drug effectiveness evaluation and disease diagnosis research of diabetic nephropathy based on quantitative techniques. First, a normal phase HPLC-ESI-TOF/MS method was used to determine the concentrations of eight phospholipid potential biomarkers during the treatment with Tangshen Formula. Results pointed out that when compared to single conventional western medicnie treatment, this traditional Chinese formula showed superiority in improving the phospholipid metabolism disorder in patients with diabetic nephropathy. However, the improvement and the onset time was not the same among all the potential biomarkers. Secondly, due to the large number of biomarkers and nonlinearity of data, a stable biomarker screening and evluation method was developed based on artificial intelligence technology.12variables were selected and the predictive accuracy was94.53%. In addition, these variables could also be used in the effectiveness evaluation of Tangshen Formula. Therefore, this method could be an alternative choice of data mining in disease diagnosis and drug effectiveness evaluation.2. Development of integrated biomarker system and its application in diabetic nephropathy. The present method of diagnosing diabetic nephropathy mainly depends on the determination of proteinuria, however, it’s very hard to perform the test in the early stage of this disease (stage1and2) for patients in this stage don’t have persistent proteinuria. According to this, an integrated biomarker system incorporating six clinical biochemical indicators, forty-one metabolic potential biomarkers from quantitative metaoblomics and five key genes was estalished and was considered effective with a high total predictive accuracy (98.9%). Then significance test and ROC analysis were employed to select the most informative potential biomarkers. Inosine with a range of0.086-0.162mg/L combined with estimated glomerular filtration rate was selected to be a sufficient potential biomarker for early diagnosis. And the levels of inosine, adenosine, S-adenosylhomocysteine and linoleic acid showed significant changes during the process of diabetic nephropathy and may be used as possible evaluation indicators for future combination therapy.3. Application of quantitative metabonomics in drug effectiveness evaluaiton of combination therapy. Promethazine is a effective drug against motion sickness, however, the sedative effects of promethazine can induce the decrease of cognitive funciton. In order to overcome this side effect, a combination therapy including caffeine and promethazine was proposed in this study. After the combination, the reduced cognitive function was improved. Then an HPLC-ESI-MS method was established to determine these two drugs and their metabolites simutaneously in all the subjects. The metabolism of promethazine was accelerated when combinaed with caffeine, and the concentraion ratio of promethazine to promethazine sulfoxide, a index of the metabolic capability, was significantly correlated with caffine and its metaoblites, which might be one of the reasons why cognitive function was improved after the combination. Besides, an OPLS-DA model was developed according to the metabonomic data, and3neurotransmitter metabolites involving dopamine and epinephrine synthesis were found significantly elevated after using caffine. Therefore, the improvement of decreased cogintive function induced by sedative effects of promethazine in combination therapy might be related with the changed endogenous metabolism including dopamine-, and epinephrine-associated pathways. |