Oxford University Press, Genetics, 4(196), p. 1227-1230, 2014
DOI: 10.1534/genetics.113.158808
Full text: Download
Abstract Exact computational methods for inference in population genetics are intuitively preferable to approximate analyses. We reconcile two starkly different estimates of the reproductive number of tuberculosis from previous studies that used the same genotyping data and underlying model. This demonstrates the value of approximate analyses in validating exact methods.