I'm performing an NMDS analysis on a Jaccard distance matrix of 347 variables across 341 subjects. I'm using mMDS from the Vegan package in R, everything seems to run fine, I get convergence and a decent stress value. When I plot the results I can see a clearly separated cluster of subjects, I have tried colouring and shaping the dots in the graph according to different variables but I don't seem to be able to find anything that explains that grouping. Is there any practical way to predict which variables might be implied in that separation (or maybe the overlapping effect of two)? Sorry if the question is basic, I'm new-ish in this.
Here are some figures to illustrate what I mean:
Many Many thanks!
