基于转录组数据的肿瘤恶性程度相关的体液标志物预测研究

The early diagnosis of cancer is very essential for control and prevention of cancer. Cancer biomarkers play an important role in early diagnosis, and also in the research on tumorigenesis and development of cancer. A biomarker is a characteristic which can be objectively measured and evaluated as an indicator of normal and cancer processes or survival rate. With the development of transcriptomics research, especially the widely used high-throughput microarray chip and RNA sequencing technology, researchers have obtained a large number of tumor related data and information on the transcriptional level. In such a mass of data and information, efficiently screening sensitive, specific and easily detected biomarkers has become one of advanced research projects of cancer biomarker research. In this project, we expect to predict sensitive, specific and easily detected cancer biomarkers which are associated with malignant degree in body fluids based on the cancer related transcriptomics data. Firstly, we collect gene and miRNA transcriptomics data of cancer from public database. Then, we develop methods and models to select tumor highly-related genes, miRNAs and identify efficient, specific cancer biomarkers. At the same time, we detect biomarkers associated with malignant degree of cancer, and study the biological mechanisms for further analysis. We will also research and develop the models to predict the protein, miRNA molecules which can be secreted into blood, urine and saliva. We could obtain the cancer biomarkers which can be detected in body fluids. Then, we try to validate the predicted cancer biomarkers in clinic by using the samples of body fluids of clinical patients. The research results of the project will be of important significance for several areas about cancer studies, such as early diagnosis and research of biological mechanisms. The ideas and methods of the project will also promote the development of bioinformatics and systems biology.

寻找有效的肿瘤诊断方法对于降低肿瘤死亡率具有重要的意义，肿瘤标志物在肿瘤早期诊断及发生发展机理研究中具有重要作用。随着转录组技术的发展，利用生物信息学技术，在海量数据中挖掘高敏感度、特异性且临床易检测的标志物，已成为当前肿瘤标志物研究的前沿课题。本项目拟在与肿瘤相关的转录组数据中，通过研发相关的新算法，筛选肿瘤高度相关的基因和miRNA，寻找高效的特异性肿瘤标志物；进一步在转录组数据中寻找与肿瘤恶性程度相关的标志物，分析肿瘤恶性程度相关的生物学机理；研发预测模型，对能够进入血液、尿液及唾液的蛋白质和miRNA分子进行预测，最终得到可在体液中检测的肿瘤恶性程度相关的标志物。在以上研究的基础上，使用临床病人的体液样本对所预测的肿瘤标志物进行临床验证。研究成果对于肿瘤早期诊断、发病机理研究等问题具有重要的意义，研究思路与方法也将为生物信息学和系统生物学等相关领域的深入研究起到积极的促进作用。

寻找有效的肿瘤诊断方法对于降低肿瘤死亡率具有重要的意义，肿瘤标志物在肿瘤早期诊断及发生发展机理研究中具有重要作用。随着转录组技术的发展，利用生物信息学技术，在海量数据中挖掘高敏感度、特异性且临床易检测的标志物，已成为当前肿瘤标志物研究的前沿课题。本项目在与肿瘤相关的转录组数据中，通过研发相关的新算法，筛选肿瘤高度相关的基因和miRNA，寻找高效的特异性肿瘤标志物；并进一步在转录组数据中寻找与肿瘤恶性程度相关的标志物，分析肿瘤恶性程度相关的生物学机理；研发预测模型，对能够进入血液、尿液及唾液的蛋白质和miRNA分子进行预测，最终得到可在体液中检测的肿瘤恶性程度相关的标志物。. 根据研究目标完成了以下工作：1) 对已有转录组数据的肿瘤，预测与分型、分期和分化度相关的基因和miRNA，得到了同种肿瘤与恶性程度相关的基因和miRNA。对不同肿瘤进行分析，得到了不同肿瘤中与恶性程度相关的基因和miRNA。2) 对已有转录组数据的肿瘤进行分子标志物预测，得到了其基因和miRNA标志物，并通过与其它肿瘤和疾病的对比研究，给出了针对特定肿瘤的特异性分子标志物，之后通过蛋白质和miRNA进入血液、尿液和唾液的预测模型获得了体液中易检测的分子标志物。 3) 针对肝癌、胃癌、结直肠癌和胰腺癌等消化系统肿瘤，使用已有血液、尿液或唾液中所得到的数据对预测得到的肿瘤标志物进行验证，并同已有的肿瘤分子标志物进行了肿瘤筛查效果的比较分析。研究成果对于肿瘤早期诊断、发病机理研究等问题具有重要的意义，研究思路和方法也将为生物信息学、系统生物学等相关领域的深入研究起到积极的促进作用。