| Metabolic reprogramming is considered to be an important hallmark of the development of many diseases.Analysis of abnormal metabolism of diseased cells can help us understand the mechanism of disease pathogenesis.Metabolic networks explain the metabolic state of cells from a systematic perspective,which is conducive to a more comprehensive analysis of the metabolic characteristics of diseases and provides more effective theoretical support for treatment.However,most studies on disease metabolism focus on the total amount of metabolism in the disease metabolic system,or on the metabolic differences in one or several specific metabolic pathways,and rarely pay attention to the changes that occur at the level of the metabolic system.Analyzing the metabolic differences of diseases from a systematic perspective will help researchers understand the mechanism of disease pathogenesis,guide the discovery of new drug targets and the design of new therapies to inhibit disease progression.In this thesis,a metabolic analysis-based method for disease drug target prediction is proposed by combining differential expression analysis and genome-scale metabolic network,which restore the changes of diseases at the level of metabolic systems,analyze metabolic characteristics,and predict drug targets.Here,the method is applied to cancer and Corona Virus Disease 2019(COVID-19).The main works include:(1)We design and implement a pan-cancer target prediction method based on multiple cancer gene expression data.The method reconstructs specific metabolic networks of multiple cancers based on cancer gene expression data,and discovers potential drug targets through gene necessity and cytotoxicity tests.By combining the results of the analysis of multiple cancers,common metabolic properties and drug targets are discovered.The results can help to understand the commonalities of metabolic system changes in different cancers and design pan-cancer therapeutic strategies.(2)We design and implement a personalized cancer target prediction method based on individual data.The method constructs personalized metabolic networks by processing the gene expression data of cancer cells and adjacent normal cells of different cancer patients,reflecting the changes of different patients at the level of the metabolic system.Some of the discovered personalized targets are consistent with existing knowledge,and some offer clues for the design of future personalized cancer treatments.(3)We design and implement a target prediction method for COVID-19 and its sequelae based on host gene expression data.The method restores changes at the metabolic system level of host cells based on gene expression data from Severe Acute Respiratory Syndrome Coronavirus-2(SARSCoV-2)infected host cell samples and normal samples.Potential metabolic drug targets are predicted by analyzing the effect of gene knockout on the synthesis of virus-related substances in host cells.The predicted drug targets have the potential to inhibit the abnormal metabolism of COVID-19,which can provide support and clues for related research. |
