申请人近期作为共同第一作者发表于Genome Biology的工作，在国际上首次实现了不依赖于遗传信息和癌旁样品，仅基于肿瘤样品的DNA甲基化数据即可推断其中肿瘤细胞的纯度及其相对于正常细胞的特异性甲基化区域。在此基础上，申请人拟引入新的研究工具和思路，建立适用性更广的单样品DNA甲基化多组分异质性分析平台，并基于此平台开展多能干细胞分化潜力和无创产前的评估研究。初步结果显示，1）平台给出的量化LOI (Loss of imprinting) 分数可以更好地评估不同来源多能干细胞的分化潜力；2）平台能够检测到孕妇外周血游离DNA中四种主要组分的一些特殊DNA甲基化标记。接下来申请人希望能够产生更多样品的DNA甲基化数据，辅以实验来证实以上结果的可靠性，同时不断通过已有数据修正算法和参数。希望我们的研究能够为干细胞重编程、无创产前检测，乃至未来癌症检测等领域提供新的研究工具、思路和参考。

The applicant’s recent study, which was published in Genome Biology, for the first time achieved tumor cell purity inferring and tumor specific methylated regions detection based on only tumor sample DNA methylation data (without genetic information and its paired normal samples). On the basis of the previous study, the applicant intends to introduce new research tools and ideas to establish a single sample DNA methylation based multicomponent heterogeneity analyzing platform for more broadly applicability. Moreover, the applicant planned to carry out pluripotent stem cell differentiation potential and noninvasive prenatal evaluation study on the platform. Preliminary results show that 1) the quantified LOI (loss of imprinting) scores provided by the platform can better predict the differentiation potential from different source of pluripotent stem cells; 2) the platform can detect some specific DNA methylation markers of four major components in maternal peripheral blood cell-free DNA. Next, the applicant hopes to generate more samples of DNA methylation high-throughput data, more experimental validations, and constantly correct the algorithms and parameters by existing data. We hope that our research can provide new research tools, ideas and references for the field of stem cell reprogramming, non-invasive prenatal testing, and the future of cancer detection.