minisv

结合泛基因组和从头个人基因组组装,提高长读长体细胞结构变异检测的灵敏度与精确度。

该项目发展长读长体细胞结构变异(SV)检测框架,结合 pangenome graphsde novo personal genome assembly,相比仅依赖参考基因组的流程提高灵敏度和精确度 (Qin et al., 2025)

肿瘤基因组在重复、高变异和低可比对区域中包含大量 SV,而这些区域常常导致参考基因组流程系统性漏检。通过组装每位个体的基因组,并在 pangenome-aware 坐标下比较肿瘤和匹配正常样本 reads,该方法能够恢复复杂体细胞事件,包括大型 indels、mobile element insertions 和 tandem duplications。源码见 GitHub。相关工作还包括基于 co-assembly graphs 的 long-range somatic SV caller colorSV (Le et al., 2025),以及从长读长数据联合检测和定相 small、structural 和 mosaic variants 的 LongcallD (Gao et al., 2026)

参考文献

2026

  1. bioRxiv
    LongcallD: joint calling and phasing of small, structural and mosaic variants from long reads
    Yan Gao, Wen-Wei Liao, Qian Qin, and 2 more authors
    bioRxiv, Mar 2026

2025

  1. bioRxiv
    Improving long-read somatic structural variant calling with pangenome and de novo personal genome assembly
    Qian Qin, Jakob Heinz, and Heng Li
    bioRxiv, Oct 2025
  2. GPB
    colorSV: Long-range Somatic Structural Variation Calling from Matched Tumor-normal Co-assembly Graphs
    M K Le, Qian Qin, and Heng Li
    Genomics, Proteomics & Bioinformatics, Sep 2025