Fig 6. True and inferred cluster size distributions at high mutation levels (×4) for each of the methods for geometrically distributed simulated cluster sizes with various means.

Results are the mean of three simulated samples with 1000 sequences each.