Run nucmer clustering without reference?

[Aciole-David]

Hi,
When performing clustering of a single file with nucmer, should I use itself as reference?

Edit: I'm trying to follow this (from https://peerj.com/articles/3817/):
Contigs from all samples were clustered with nucmer (Delcher, Salzberg & Phillippy, 2003) at ≥95% ANI across ≥80% of their lengths, as in (Brum et al., 2015; Gregory et al., 2016), to generate a pool of non-redundant “population contigs”.

obs: ANI=average nucleotide identity

What are the nucmer parameters to set in order to use the values mentioned?
Thank you