There is statistics and then there is statistics. For example this paper is statistics based https://www.nature.com/articles/ng.3487 with some novel data too. But its stats over graphs, which is way less straight forward. You don't just fire up R and do model <- lme(data, x ~ y + z) That paper spawned a fairly amazing startup too one that I really really really doubt anyone could possibly have come up with without a PhD and a lot of insight https://mogrify.co.uk/