Fig. 1

Overview of the CloneCNA probabilistic framework. a CloneCNA analysis workflow. Three inputs are required: 1) exon-level read counts of paired tumor-normal samples; 2) tumor allelic read depths of germline heterozygous SNP positions; 3) GC-content of all exon regions. The LCR and MAF data is modeled using a factorial HMM. The number of clonal clusters is determined using BIC. b HMM adopted in CloneCNA. Two Markov chains are adopted to delineate copy number aberrations and clonal clusters