MiScan: mutation-inferred screening model of breast cancerΒΆ
Here we present MiScan: mutation-inferred screening model of cancer for screening of breast cancer patients using genotype information from blood samples. We coarsened 119,318 variants into 13,885 genomic windows, and trained MiScan with coarse-grained mutation information from 1183 breast cancer patients and 3232 healthy people. To promote the application of MiScan for breast cancer prediction and prevention, we established an interactive website, which automatically provides a breast cancer risk assessment feedback report from genotype information as input.