近年来癌症已经成为我国居民死亡的首要原因，基因组变异被认为是大多数癌症的发病原.因。这些变异涉及大量高频的基因组重组、多拷贝数变异、DNA核苷酸变异及表观遗传变异等复杂的结构和功能进化事件，因此从分子水平研究癌症发生和发展过程，是癌症预防，检测及治疗的关键途径。近些年来高通量测序技术的发展使对不同时期癌症全基因组测序成为可能。传统的生物学方法只能对癌基因组进行局部和静态研究。因此，从进化的角度，在全基因组的水平上研究癌症的变异过程和进化历史有着非常重要的意义。本项目融合数学，统计和计算优化，基于高通量测序数据，对癌细胞基因组进化，重构，变异和进化等问题，开发准确的数学模型和高效的算法。通过对不同时期癌细胞基因全基因组测序数据进行比对和分析，重建不同时期癌细胞基因组之间进化关系和变异过程，结合基因组结构和基因功能改变，从而为医学基因水平诊断，精确治疗和药物开发提供基础。

Cancer, a leading cause of death in China, is a disease caused by genomic variations. These genomic variations include genome rearrangements, chromothripsis, copy number variation(CNV), structural variations (SV), simple single nucleotide variations (SNV) and epigenetic changes. Detailed characterization of these genomic variations can provide crucial insight about tumorigenesis and for cancer treatment. The advanced high-throughput sequencing technologies (HTS) has offered us an opportunity to sequence the whole genome data of cancer genomes from different periods. However, current biological methods only allow for local and static characterization of genomic variations in tumor genomes. Thus it is crucial to utilize whole genome information to fully understand cancer structural variations and evolutionary history. This project combines methods from mathematics, statistics and computer science, with the focus on modeling and analyzing cancer genomic evolution, genome reconstruction, structural variations and evolution. We will develop robust mathematical models and computational tools to compare whole genome sequencing data from various stages of tumor so that we can establish cancer genome evolutionary path and corresponding functional variation on tumorigenesis, which has important implications for cancer diagnosis, treatment and medicine design.