Amplification of DNA mixtures--Missing data approach

Torben Tvedebrink; Poul Svante Eriksen; Helle Smidt Mogensen; Niels Morling

doi:10.1016/j.fsigss.2007.08.002

Amplification of DNA mixtures--Missing data approach

Torben Tvedebrink, Poul Svante Eriksen, Helle Smidt Mogensen, Niels Morling

Section of Forensic Genetics

5 Citations (Scopus)

Abstract

This paper presents a model for the interpretation of results of STR typing of DNA mixtures based on a multivariate normal distribution of peak areas. From previous analyses of controlled experiments with mixed DNA samples, we exploit the linear relationship between peak heights and peak areas, and the linear relations of the means and variances of the measurements. Furthermore, the contribution from one individual's allele to the mean area of this allele is assumed proportional to the average of height measurements on alleles where the individual is the only contributor. For shared alleles in mixed DNA samples, it is only possible to observe the cumulative peak heights and areas. Complying with this latent structure, we use the EM-algorithm to impute the missing variables based on a compound symmetry model. That is the measurements are subject to intra- and inter-loci correlations not depending on the actual alleles of the DNA profiles. Due to factorization of the likelihood, properties of the normal distribution and use of auxiliary variables, an ordinary implementation of the EM-algorithm solves the missing data problem. We estimate the parameters in the model based on a training data set. In order to assess the weight of evidence provided by the model, we use the model with the estimated parameters on STR data from real crime cases with DNA mixtures

Original language	English
Title of host publication	Progress in Forensic Genetics 12 : Proceedings of the 22nd International ISFG Congress
Editors	Niels Morling
Volume	1
Publisher	Elsevier
Publication date	2008
Pages	664-666
DOIs	https://doi.org/10.1016/j.fsigss.2007.08.002
Publication status	Published - 2008
Event	22nd International ISFG Congress - Copenhagen, Denmark Duration: 21 Aug 2007 → 25 Aug 2007

Conference

Conference	22nd International ISFG Congress
Country/Territory	Denmark
City	Copenhagen
Period	21/08/2007 → 25/08/2007

Series	Forensic Science International: Genetics Supplement Series
Number	1
Volume	1
ISSN	1875-1768

Access to Document

10.1016/j.fsigss.2007.08.002

Cite this

Tvedebrink, T., Eriksen, P. S., Mogensen, H. S., & Morling, N. (2008). Amplification of DNA mixtures--Missing data approach. In N. Morling (Ed.), Progress in Forensic Genetics 12: Proceedings of the 22nd International ISFG Congress (Vol. 1, pp. 664-666). Elsevier. Forensic Science International: Genetics Supplement Series Vol. 1 No. 1 https://doi.org/10.1016/j.fsigss.2007.08.002

Amplification of DNA mixtures--Missing data approach. / Tvedebrink, Torben; Eriksen, Poul Svante; Mogensen, Helle Smidt et al.

Progress in Forensic Genetics 12: Proceedings of the 22nd International ISFG Congress. ed. / Niels Morling. Vol. 1 Elsevier, 2008. p. 664-666 (Forensic Science International: Genetics Supplement Series; No. 1, Vol. 1).

Research output: Chapter in Book/Report/Conference proceeding › Article in proceedings › Research › peer-review

Tvedebrink, T, Eriksen, PS, Mogensen, HS & Morling, N 2008, Amplification of DNA mixtures--Missing data approach. in N Morling (ed.), Progress in Forensic Genetics 12: Proceedings of the 22nd International ISFG Congress. vol. 1, Elsevier, Forensic Science International: Genetics Supplement Series, no. 1, vol. 1, pp. 664-666, 22nd International ISFG Congress, Copenhagen, Denmark, 21/08/2007. https://doi.org/10.1016/j.fsigss.2007.08.002

@inproceedings{1dc20d90a20011debc73000ea68e967b,

title = "Amplification of DNA mixtures--Missing data approach",

abstract = "This paper presents a model for the interpretation of results of STR typing of DNA mixtures based on a multivariate normal distribution of peak areas. From previous analyses of controlled experiments with mixed DNA samples, we exploit the linear relationship between peak heights and peak areas, and the linear relations of the means and variances of the measurements. Furthermore, the contribution from one individual's allele to the mean area of this allele is assumed proportional to the average of height measurements on alleles where the individual is the only contributor. For shared alleles in mixed DNA samples, it is only possible to observe the cumulative peak heights and areas. Complying with this latent structure, we use the EM-algorithm to impute the missing variables based on a compound symmetry model. That is the measurements are subject to intra- and inter-loci correlations not depending on the actual alleles of the DNA profiles. Due to factorization of the likelihood, properties of the normal distribution and use of auxiliary variables, an ordinary implementation of the EM-algorithm solves the missing data problem. We estimate the parameters in the model based on a training data set. In order to assess the weight of evidence provided by the model, we use the model with the estimated parameters on STR data from real crime cases with DNA mixtures",

author = "Torben Tvedebrink and Eriksen, {Poul Svante} and Mogensen, {Helle Smidt} and Niels Morling",

year = "2008",

doi = "10.1016/j.fsigss.2007.08.002",

language = "English",

volume = "1",

series = "Forensic Science International: Genetics Supplement Series",

publisher = "Elsevier",

number = "1",

pages = "664--666",

editor = "Niels Morling",

booktitle = "Progress in Forensic Genetics 12",

address = "Netherlands",

note = "22nd International ISFG Congress ; Conference date: 21-08-2007 Through 25-08-2007",

}

TY - GEN

T1 - Amplification of DNA mixtures--Missing data approach

AU - Tvedebrink, Torben

AU - Eriksen, Poul Svante

AU - Mogensen, Helle Smidt

AU - Morling, Niels

PY - 2008

Y1 - 2008

N2 - This paper presents a model for the interpretation of results of STR typing of DNA mixtures based on a multivariate normal distribution of peak areas. From previous analyses of controlled experiments with mixed DNA samples, we exploit the linear relationship between peak heights and peak areas, and the linear relations of the means and variances of the measurements. Furthermore, the contribution from one individual's allele to the mean area of this allele is assumed proportional to the average of height measurements on alleles where the individual is the only contributor. For shared alleles in mixed DNA samples, it is only possible to observe the cumulative peak heights and areas. Complying with this latent structure, we use the EM-algorithm to impute the missing variables based on a compound symmetry model. That is the measurements are subject to intra- and inter-loci correlations not depending on the actual alleles of the DNA profiles. Due to factorization of the likelihood, properties of the normal distribution and use of auxiliary variables, an ordinary implementation of the EM-algorithm solves the missing data problem. We estimate the parameters in the model based on a training data set. In order to assess the weight of evidence provided by the model, we use the model with the estimated parameters on STR data from real crime cases with DNA mixtures

AB - This paper presents a model for the interpretation of results of STR typing of DNA mixtures based on a multivariate normal distribution of peak areas. From previous analyses of controlled experiments with mixed DNA samples, we exploit the linear relationship between peak heights and peak areas, and the linear relations of the means and variances of the measurements. Furthermore, the contribution from one individual's allele to the mean area of this allele is assumed proportional to the average of height measurements on alleles where the individual is the only contributor. For shared alleles in mixed DNA samples, it is only possible to observe the cumulative peak heights and areas. Complying with this latent structure, we use the EM-algorithm to impute the missing variables based on a compound symmetry model. That is the measurements are subject to intra- and inter-loci correlations not depending on the actual alleles of the DNA profiles. Due to factorization of the likelihood, properties of the normal distribution and use of auxiliary variables, an ordinary implementation of the EM-algorithm solves the missing data problem. We estimate the parameters in the model based on a training data set. In order to assess the weight of evidence provided by the model, we use the model with the estimated parameters on STR data from real crime cases with DNA mixtures

U2 - 10.1016/j.fsigss.2007.08.002

DO - 10.1016/j.fsigss.2007.08.002

M3 - Article in proceedings

VL - 1

T3 - Forensic Science International: Genetics Supplement Series

SP - 664

EP - 666

BT - Progress in Forensic Genetics 12

A2 - Morling, Niels

PB - Elsevier

T2 - 22nd International ISFG Congress

Y2 - 21 August 2007 through 25 August 2007

ER -