Fig. 1.

Mapping between two reads. $b[1]$ and $e[1]$ are the 0-based starting and ending mapping coordinates of the first read v, respectively. $b[2]$ and $e[2]$ are the mapping coordinates of read w. Lightgray areas indicate overhang regions that should be mapped together if the overlap is real. If the overhang regions are small enough, the figure implies an edge $v\to w$ with approximate length $\ell (v\to w)=b[1]-b[2]$ and its complement edge $\overline{w}\to \overline{v}$ with $\ell (\overline{w}\to \overline{v})=(l[2]-e[2])-(l[1]-e[1])$