And second of all, it has to have length at least 4.

So this is just for simplicity.

We could just as well have allowed the overlaps to have mismatches and

gaps in them, but here for simplicity we don't.

And we want to make them have length at least 4 to reduce the chance that we get

a not very convincing overlap between two of these reads.

So given that threshold, we can write out the entire overlap graph,

which is what's shown at the bottom of this slide here.

And here the nodes here correspond to reads, and

they're each labeled accordingly, and the directed edges correspond to the overlaps.

And here the edges are also labeled.

Each edge is labeled with the length of the corresponding suffix-prefix match.

So for example, this edge is labeled with a 5 because a length 5 suffix,

ACGTA, matches a length 5 prefix, ACGTA, here.