Minimal overlap between two reads in the de Bruijn algorithm

[TauOvermind]

I was reading a paper named "Comparison of the two major classes of assembly algorithms: overlap–layout–consensus and de-bruijn-graph" by Zhenyu Li et al. (2011), when I came across the following sentence:

In the DBG algorithm, to ensure the k-mers can be linked, the minimum overlap between two reads should be no less than the k-mer size K that is equivalent to the T parameter in the Lander–Waterman model.

I thought that the minimum overlap between two reads should be no less than K-1 and not K. I would really appreciate it if someone could explain me the quoted statement.

[Brian Bushnell]

I thought that the minimum overlap between two reads should be no less than K-1 and not K. I would really appreciate it if someone could explain me the quoted statement.

Yes, you're right. In practice there's so little difference between K and K-1 that it's easy to make a mistake like that and overlook it.

[TauOvermind]

Thank you, Brian. I agree that the difference between these two values would be minimal in practice, but as it was a paper about the theory of sequence assembly, I found that statement confusing, as I thought that I might have been missing an important point in the derivation.