Figure 1

Reduce memory footprint by genome partitioning and analysis. (a) The reference genome, denoted by a horizontal line, is partitioned into 6 non-overlapping windows W _i s (0≤i≤5). (b) When analyzing each window, all reads overlapping this window as well as the relevant paired end reads are considered. Each read (denoted by a short line) is placed underneath the location where it is aligned to the reference. Each read pair is connected by a dotted line. Assuming W₁ is the target window, all reads overlapping with W₁ will be considered for pileup and phase analysis. In addition, reads overlapping with W_1is, 1≤i≤3, will be considered for phase analysis.