The paper was presented to and discussed at the RSS at 11 June 2014 and is published with contributed discussions as well as the authors' response to the discussion.

• The paper presents a statistical model for DNA mixtures; DNAmixtures implements this model.
• All analyses are performed using DNAmixtures and are documented as a short tutorial in the supporting information
  Case analysis using the DNAmixtures package by Therese Graversen.

The paper introduces generally applicable methodology for

• efficient exact summation over unknown genotypes.
• efficient exact computation of the likelihood function.
• easy evaluation of various marginal and conditional distributions of peak heights.
• statistical methodology for systematic scrutiny of DNA mixture models.

This Doctoral thesis is freely available from the Oxford Bodleian Libraries and

• presents an in-depth discussion of the methodology.
• illustrates how a statistical framework may be exploited to address a wide range of questions relating to the analysis of a DNA mixture.
• gives an overview of the DNAmixtures package.