A Bayesian model integration for mutation calling through data partitioning

• DOI: 10.1093/bioinformatics/btz233
• Publication Date: 2019-03-28T12:54:42Z
• Authors: Takuya Moriyama; Seiya Imoto; Shuto Hayashi; Yuichi Shiraishi; Satoru Miyano; Rui Yamaguchi
• Keywords: Generative model

Detection of somatic mutations from tumor and matched normal sequencing data has become among the most important analysis methods in cancer research. Some existing mutation callers have focused on additional information, e.g. heterozygous single-nucleotide polymorphisms (SNPs) nearby candidates or overlapping paired-end read information. However, cannot take multiple information sources into account simultaneously. Existing Bayesian hierarchical model-based construct two generative models, model error model, limited been modeled.We proposed a integration framework named as partitioning-based integration. In this framework, through introducing partitions for reads based given sources, we integrate models utilize sources. Based that, constructed novel method OHVarfinDer. both introduced set that cover candidate position, applied different each category reads. We demonstrated our can SNP effectively simulation datasets real datasets.https://github.com/takumorizo/OHVarfinDer.Supplementary are available at Bioinformatics online.