a Department of Biostatistics, Mailman School of Public Health , Columbia University , New York , New York , U.S.A.
b The Jackson Laboratory for Genomic Medicine , Farmington , Connecticut , U.S.A.
Epigenetics. 2018;13(2):192-206. doi: 10.1080/15592294.2017.1389361. Epub 2018 Feb 6.
Cell division is important in human aging and cancer. The estimation of the number of cell divisions (mitotic age) of a given tissue type in individuals is of great interest as it allows not only the study of biological aging (using a new molecular aging target) but also the stratification of prospective cancer risk. Here, we introduce the MiAge Calculator, a mitotic age calculator based on a novel statistical framework, the MiAge model. MiAge is designed to quantitatively estimate mitotic age (total number of lifetime cell divisions) of a tissue using the stochastic replication errors accumulated in the epigenetic inheritance process during cell divisions. With the MiAge model, the MiAge Calculator was built using the training data of DNA methylation measures of 4,020 tumor and adjacent normal tissue samples from eight TCGA cancer types and was tested using the testing data of DNA methylation measures of 2,221 tumor and adjacent normal tissue samples of five other TCGA cancer types. We showed that within each of the thirteen cancer types studied, the estimated mitotic age is universally accelerated in tumor tissues compared to adjacent normal tissues. Across the thirteen cancer types, we showed that worse cancer survivals are associated with more accelerated mitotic age in tumor tissues. Importantly, we demonstrated the utility of mitotic age by showing that the integration of mitotic age and clinical information leads to improved survival prediction in six out of the thirteen cancer types studied. The MiAge Calculator is available at http://www.columbia.edu/∼sw2206/softwares.htm .
细胞分裂在人类衰老和癌症中很重要。估计个体中特定组织类型的细胞分裂次数（有丝分裂年龄）非常有趣，因为它不仅允许研究生物衰老（使用新的分子衰老靶标），还允许分层预测癌症风险。在这里，我们介绍了 MiAge 计算器，这是一种基于新统计框架的有丝分裂年龄计算器，即 MiAge 模型。MiAge 旨在使用在细胞分裂过程中表观遗传遗传过程中积累的随机复制错误来定量估计组织的有丝分裂年龄（一生中的细胞分裂总数）。使用 MiAge 模型，根据来自八个 TCGA 癌症类型的 4020 个肿瘤和相邻正常组织样本的 DNA 甲基化测量的训练数据以及另外五个 TCGA 癌症类型的 2221 个肿瘤和相邻正常组织样本的 DNA 甲基化测量的测试数据，构建了 MiAge 计算器。我们表明，在所研究的十三种癌症类型中的每一种中，肿瘤组织中的有丝分裂年龄普遍比相邻正常组织加速。在这十三种癌症类型中，我们表明肿瘤组织中更快的有丝分裂年龄与更差的癌症存活率相关。重要的是，我们通过表明将有丝分裂年龄与临床信息相结合可以改善六种研究的癌症类型的生存预测，证明了有丝分裂年龄的实用性。MiAge 计算器可在 http://www.columbia.edu/∼sw2206/softwares.htm 获得。
