Polimorfismos do tipo indel são os mais abundantes depois dos SNPs, representando 3,6 milhões das variantes caracterizadas pelo projeto 1000 Genomes. Com uma distribuição que pode ser estimada em mais de um indel a cada 1000 pb, são facilmente encontrados em regiões de interesse. A baixa taxa de mutação e a possibilidade de desenhar primers alelo-específicos são as principais características dos indels que os diferenciam de STRs. O uso de primers aleloespecíficos na detecção e dosagem de misturas de DNA apresenta maior sensibilidade e acurácia que as técnicas usualmente empregadas. Aqui foram descritos, para 10 lócus indel, pares de primers flanqueadores e alelo-específicos para ambos os alelos (inserção e deleção) e foi realizado o estudo populacional em 160 indivíduos. A determinação de fenótipos e avaliação de especificidade dos primers, dos quais 28 foram específicos, foi realizada por PCR convencional seguida de PAGE. As análises populacionais e forenses mostraram que esses lócus apresentam alta variabilidade (heterozigose de 30-50%) e consequentemente, alta informatividade. Os valores de PIC, PE e PD variaram de 0,2763 a 0,3750; 0,1381 a 0,1875 e 0,4978 a 0,6250 respectivamente. Os valores cumulativos de PCE e PCD foram respectivamente 0,8508 e 0,9999. Assim, esse conjunto de indels é indicado para serem testados na detecção e quantificação de misturas de DNA a partir da amplificação alelo-específica. / Indels polymorphisms are the most abundant after SNPs, representing 3.6 million of the variants characterized by the 1000 Genomes project. With a distribution that can be estimated at more than one indel per 1000 bp, they are easily found in regions of interest. The low mutation rate and the possibility of designing allele-specific primers are the main characteristics of the indels that differentiate them from STRs. The use of allele-specific primers in the detection and dosage of DNA mixtures is more sensitive and accurate than regularly employed techniques. Here, for 10 indel loci, pairs of flanking primers and allele-specific primers, for both alleles (insertion and deletion), were described and a population study was performed on 160 individuals. Determining phenotypes and evaluation of primers specificity, of which 28 were specific, was performed by conventional PCR followed by PAGE. In population and forensic analysis, these loci showed high variability (heterozygosis of 30-50%) and consequently high informativeness. The values of PIC, PE and PD ranged from 0.2763 to 0.3750, 0.1381 to 0.1875 and 0.49978 to 0.6250 respectively. Combined values of PCE and PCD were respectively 0.8508 and 0.9999. Thus, this set of indels is indicated to be tested for detection and quantification of DNA mixtures using the allele-specific amplification method.

L’analyse et l’interprétation d’´echantillons constitu´es de mélanges d’ADN de plusieurs individus est un défi majeur en science forensique. Lorsqu’un expert de la police scientifique a affaire à un mélange d’ADN il doit répondre à deux questions: d’abord, “combien de contributeurs y a-t-il dans ce mélange ?”et puis, “quels sont les génotypes des individus impliqués ?” Le typage seul de cet ADN ne permet pas toujours de r´epondre `a ces questions. En effet leproblème est posé d`es lors que plus de deux allèles sont observées à un locus donné, plusieurscombinaisons génotypiques sont alors `a envisager et il est impossible de déterminer avec certitudele nombre d’individus qui ont contribué au m´elange. De plus, la présence d’anomalies liées àl’analyse de marqueurs g´en´etiques, comme la contamination ou la perte d’all`eles (“drop-out”),peut davantage compliquer l’analyse.Les nombreux d´eveloppements statistiques d´edi´es `a ces probl´ematiques n’ont pas eu le succ`esescompt´e dans la communaut´e forensique, essentiellement, parce que ces m´ethodes n’ont pas ´et´evalid´ees. Or sans cette validation, les experts de la police scientifique ne peuvent exploiter cesm´ethodes sur des m´elanges issus d’affaires en cours d’investigation.Avant d’ˆetre valid´ees, ces m´ethodes doivent passer par une rigoureuse ´etape d’´evaluation.Cette derni`ere soul`eve deux questions: d’abord, la question de la m´ethodologie `a adopter, puis,celle des outils `a d´eployer. Dans cette th`ese, nous tentons de r´epondre aux deux questions.D’abord, nous menons des ´etudes d’´evaluation sur des m´ethodes d´edi´ees `a deux questions cl´es: i)l’estimation du nombre de contributeurs `a un m´elange d’ADN et ii) l’estimation des probabilit´esde “drop-out”. En second lieu, nous proposons un logiciel “open-source” qui offre un certainnombre de fonctionnalit´es permettant de faciliter l’´evaluation de m´ethodes statistiques d´edi´eesaux m´elanges d’ADN.Cette thèse a pour but d’apporter une r´eponse concr`ete aux experts de la police scientifiqueen leur fournissant `a la fois une d´emarche m´ethodologique pour l’´evaluation de m´ethodes, et lapossibilit´e d’analyser la sensibilit´e de leurs r´esultats au travers d’un outil informatique en libreacc`es. / Analysis of forensic DNA mixtures recovered from crime scenes is one of the most challengingtasks in forensic science. DNA mixture raise two main questions: “how many contributors arethere” and “what are the genotypes of the contributing individuals?” The genetic characterizationalone of such samples does not always answer these questions. In fact, whenever more thantwo alleles are observed at a given locus, several distinct genotypic combinations are plausiblefor the unknown contributors to the sample, and it is not possible to determine the number ofthese contributors with absolute certainty. Besides, the presence of anomalies related to DNAtyping techniques, such as contamination or allele loss (drop-out), can further complicate theanalysis.Numerous statistical developments facilitating DNA mixtures interpretation were proposed,but they did not receive the expected success in the forensic community. The main explanationfor this is that these methods are not validated for forensic casework.In order to achieve this validation criterion, the methods must undergo a rigorous evaluationstep. The latter raises two questions: i) how methods should be evaluated? and ii) whattools can be used to conduct evaluation studies? In this thesis we attempt to answer bothquestions. First, we evaluate methods dedicated to two key issues, the estimation of the numberof contributors to DNA mixtures and the estimation of drop-out probabilities. Second, wepropose an “open-source” software that offers a number of functionalities dedicated to facilitatingmethod evaluation through the simulation of data commonly encountered in forensic settings.This thesis aims to provide a concrete answer to the issues raised by forensic DNA mixtures,by providing a methodology for method evaluation and by offering necessary tools to enablemethod evaluation.

Die Haplotypspezifische Extraktion (HSE) bietet für die Analyse von männlichen Mischprofilen einen neuen und direkteren Lösungsansatz, in dem die haploiden Y-chromosomalen DNS-Komponenten der einzelnen Individuen bereits vor der Analyse der individual spezifischen Marker separiert werden können und dadurch eine wirkliche physische Trennung erreicht wird. Die HSE verwendet Y-chromosomalen SNPs für die Erstellung allelspezifische Extraktionssonden, die nun gezielt nur die Marker der extrahierten DNS Komponente bzw. einer Person separieren sollen. Dabei werden im Hybridisationsschritt der HSE selektive nur komplett hybridisierte Sonden durch eine Polymerase verlängert. Während der Elongation erfolgt eine Biotinylierung des neu entstehenden Stranges, welcher dann selektiv durch Streptavidin markierte Eisenkügelchen extrahiert werden kann. Erste Durchführungen einer HSE zeigten nur eine sehr schwache bis keine Anreicherung. Während der Optimierung verschiedener Parameter wurde die Schlüsselstellung des Sondendesigns in der HSE-Technik deutlich. Die Ergebnisse zeigten, dass die neu entwickelten Sonden den Trennungserfolg der Mischprobe enorm verbessern und in einigen Fällen sogar zum Ausschluss des konkurrierenden Allels führten. Ein Vergleich der HSE Ergebnisse mit den simulierten Sondenparametern der getesteten Sonden ergab, dass der Extraktionserfolg der Sonde maßgeblich durch das Zusammenspiel von Sondenlänge und GC-Gehalt bestimmt wird. Durch dieses neu gewonnene Verständnis über den Einfluss der einzelnen Sondenparameter auf den Trennungserfolg der Mischprobe, können für künftige HSE Anwendungen Sonden effizienter erstellt und deren Wirksamkeit vorhergesagt werden. Zusätzlich konnte das neu entwickelte Vorhersage-Model der Sondenspezifität auch für weitere Extraktionsorte außerhalb des Y-Chromosoms bestätigt werden. Weiterhin konnte durch die Kombination verschiedener Sonden in einer Multiplex HSE mehrerer Y-chromosomaler Marker gleichzeitig getrennt. / Haplotype-specific extraction (HSE) allows the separation of diploid samples in their haploid components and offers in forensic a new straight forward method to separate Y-chromosomal mixed profiles, consisting of haplotype markers like short tandem repeats (STRs). The advantage of the HSE approach in mixture analysis is the real physical separation of the individual DNA components before the amplification of the STR markers. In order to use the HSE technique for the separation of male DNA mixtures, Y-chromosomal extraction probes were designed to single nucleotide polymorphisms (SNPs), which have been specific for one contributor of the male DNA mixtures. During extraction only complete matched probes are extended by a polymerase which results in the incorporation of biotinylated nucleotides. The synthesized and biotin labeled strand is separated by streptavidin coated magnetic beads. Finally, samples were analyzed by PCR coupled capillary electrophoresis for the detection of the extracted STR markers. First tests of a Y-chromosomal specific extraction showed only little till no enrichment of the targeted alleles. Therefore optimization tests of different parameters were carried out, which revealed the probe design as the key factor of successful HSE. A comparison between simulated probe parameters and their extraction success in HSE showed that the HSE probe efficiency mainly depends of the relation of probe length and GC-contents. Because of the new gained knowledge about the influence of the probe-design on the separation success, probes for future HSE application can be developed faster and cost-effective. The new prediction model for probe-specificity was also successful tested for the extraction of other genome-loci. Furthermore, a multiplex HSE approach was used to separate several STR markers simultaneously in one extraction reaction and therefore achieved the separation of one contributor Y-chromosomal haplotype.

This thesis proposes and discusses a statistical model for interpreting forensic DNA mixtures. We develop methods for estimation of model parameters and assessing the uncertainty of the estimated quantities. Further, we discuss how to interpret the mixture in terms of predicting the set of contributors. We emphasise the importance of challenging any interpretation of a particular mixture, and for this purpose we develop a set of diagnostic tools that can be used in assessing the adequacy of the model to the data at hand as well as in a systematic validation of the model on experimental data. An important feature of this work is that all methodology is developed entirely within the framework of the adopted model, ensuring a transparent and consistent analysis. To overcome the challenge that lies in handling the large state space for DNA profiles, we propose a representation of a genotype that exhibits a Markov structure. Further, we develop methods for efficient and exact computation in a Bayesian network. An implementation of the model and methodology is available through the R package DNAmixtures.