Identification and separation of DNA mixtures using peak area information

Abstract

We show how probabilistic expert systems can be used to analyse forensic identification problems involving DNA mixture traces using quantitative peak area information. Peak area is modelled with conditional Gaussian distributions. The expert system can be used for ascertaining whether individuals, whose profiles have been measured, have contributed to the mixture, but also to predict DNA profiles of unknown contributors by separating the mixture into its individual components. The potential of our methodology is illustrated on case data examples and compared with alternative approaches. The advantages are that identification and separation issues can be handled in a unified way within a single network model and the uncertainty associated with the analysis is quantified.Some key words and phrases: Bayesian network, conditional Gaussian distributions, DNA mixture, DNA profile, forensic identification, mixture separation, probabilistic expert system, peak weight.

Probabilistic expert systems for handling artifacts in complex DNA mixtures

“…For a detailed exposition of this model and the implementation as a Probabilistic Expert System (PES) (Cowell et al 1999) we refer the reader to Cowell et al (2006Cowell et al ( , 2007aCowell et al ( , 2007b.…”

“…Under this hypothesis, in contrast to H p , it is no longer definite that the MC18 peak of allele 22 (D2) arises as a stutter peak from allele 23; it could be that the unknown person has one or both alleles. Using methods described in Cowell et al (2007b), it is possible to use the same network for the separation of DNA profiles, and so give an estimate of the most likely DNA profile of the third contributor conditional on the peak height values and the DNA profiles of K 1 and K 3 . Such a simultaneous analysis of both mixtures yields the following posterior probability distribution that allele 23 (D2) in MC18 stutters: 0.565 (at None), 0.176 (at 5%), and 0.259 (at 10%).…”

Probabilistic characterisation of baseline noise in STR profiles

“…Interpretation methods that utilise peak height or area information, rather than the binary information of whether a peak height or area exceeds a threshold or not, have garnered much attention, and a variety of fully-continuous interpretation platforms have been developed [9][10][11][12][13][14][15][16].…”

“…The peak area information is used in [9][10][11][12] to establish a quantitative method for the interpretation of STR mixtures. In There are three dominant contributing factors that distort short tandem repeat profile measurements, two of which, stutter and variations in the allelic peak heights, have been described extensively.…”

“…Sometimes, noise is modelled implicitly as small peaks with mean zero [10,12]. In [13], a continuous modelling of the peak height is used if the peak exceeds a user defined analytical threshold.…”