Figure 1
From: A Hidden Markov Model for Copy Number Variant prediction from whole genome resequencing data
The Hidden Markov Model. The states: (1) Normal: normal diploid state, copy number 2, mate pair distance follows the empirical distribution observed from genome-wide data. (2) Del 1: 1-copy deletion state, mate pair distance follows the empirical distribution observed from genome-wide data. (3) Del 2: 0-copy deletion state, mate pair distance follows the empirical distribution observed from genome-wide data. (4) Dup 1: 3-copy duplication state, mate pair distance follows the empirical distribution observed from genome-wide data. (5) Dup 2: 4-copy duplication state. This state is also capable of capturing duplications with more than 4 copies. (6) Grid states: marked with purple dashed frames. Each row of grid states targets deletions of a particular size. The 5’ flanking and 3’ flanking states (collectively, F states) model the flanking regions of such deletions where the mate pair distances follow the empirical distribution with a shifted mean value (genome-wide average + deletion size).