Managed to open it from one of these links eventually:) It's a really nice analysis. It makes me wonder, are these data driven clusters also related to phenotype or some external measure that differentiates these individuals?

Great question. We've looked into some of this in a previous paper (www.jneurosci.org/content/42/2...) and didn't find anything glaring pointing in this direction, but we're collecting more data on what behavioral correlates might explain the individual differences.