Welcome to the BreCANDB website

About BreCANDB
BreCAN-DB is DataBase of Breakpoint profiles of CANcer genomes, a repository of personalized Cancer somatic DNA breakpoint profiles mapped using whole genome sequencing data mostly obtained from The Cancer Genome Atlas. It also provides a framework to visualize these breakpoints using a Breakpoints browser, also accessible through navigation bar. In the current release of BreCAN-DB v1.0, a total of ~1,92,500 breakpoints were mapped at nucleotide resolution in 99 cancer-normal pairs of whole genome samples using Meerkat software (Yang et. al., Cell, 2013) in five types of cancer, namely,

  • Glioblastoma multiforme (GBM) (Brennan et. al., Cell, 2013)
  • Breast invasive carcinoma (BRC) (Banerji et. al., Nature, 2012)
  • Lung adenocarcinoma (LUAD) (Collisson et. al., Nature, 2014)
  • Ovarian serous cystadenocarcinoma (OV) (TCGA network, Nature, 2011)
  • Head and neck squamous cell carcinoma (HN) (Lawrence et. al., Cell, 2015)

  • A total of 138 whole genome breakpoint profiles of ten cancer types from Yang et. al., Cell, 2013 have also been included in breakpoint browser for visualization.

    **If you use this database, please cite us as
    BreCAN-DB : A repository cum browser of personalized DNA breakpoint profiles of cancer genomes. Nucleic Acids Research, 2015 , doi: 10.1093/nar/gkv1264

    Mapping of somatic DNA breakpoints in cancer genomes

    DNA Breakpoints in cancer and corresponding normal genomes were mapped using Meerkat at default parameters. Meerkat uses cluster of discordant pairs and soft-clipped reads to map breakpoints at single base resolution. These breakpoints correspond to large-sized genomic alterations such as deletions, insertions, inversions, tandem duplications, translocations. and complex events. False postives were removed by using genomic alterations supported by atleast two discordant pairs and two clipped reads. Further, genomic alterations with size larger than 10mb in size were also removed as such events are very rare. Events overlapping with Database of Genomic Variants and pooled datasets of normal genomes of corresponding cancer type were also removed to map somatic breakpoints.
    Mapping of Cancer DNA breakpoint hotspots

    To identify tumor-specific breakpoint hotspots, we extended breakpoint of each tumor-specific genomic alteration by 10 Kb on both sides which we call as breakpoint extended regions. These extended regions have been shown as rectangular bars in breakpoint profiles tracks of browser. Taking boundaries of each extended region, we divided the whole genome into non-overlapping bins of unequal sizes. The bins that overlapped with extended region in more than one tumor genome were identified and their enrichment score (A) was calculated as number of samples harboring GAs i.e.

    Enrichment score = number of samples harboring genomic alterations

    Next, we compared these scores with expected distribution of score calculated by randomizing the GAs of all samples 10000 times and calculated their significance. The bins with p-value < 0.05 after Bonferroni hochberg correction for multiple hypothesis testing were finally selected as breakpoint hotspots (shown as rectangular bars in breakpoints browser).


    October, 2015
    DNA breakpoint profiles added for 138 pairs of samples from Yang et al., Cell 2013
    June, 2015
    DNA breakpoint profiles added for 99 pairs of samples