Forensic Practitioner's Guide to the Interpretation of Complex DNA Profiles

Over The Past Twenty Years, There's Been A Gradual Shift In The Way Forensic Scientists Approach The Evaluation Of Dna Profiling Evidence That Is Taken To Court. Many Laboratories Are Now Adopting 'probabilistic Genotyping' To Interpret Complex Dna Mixtures. However, Current Practice Is Very Diverse, Where A Whole Range Of Technologies Are Used To Interpret Dna Profiles And The Software Approaches Advocated Are Commonly Used Throughout The World. Forensic Practitioner's Guide To The Interpretation Of Complex Dna Profiles Places The Main Concepts Of Dna Profiling Into Context And Fills A Niche That Is Unoccupied In Current Literature. The Book Begins With An Introduction To Basic Forensic Genetics, Covering A Brief Historical Description Of The Development And Harmonization Of Str Markers And National Dna Databases. The Laws Of Statistics Are Described, Along With The Likelihood Ratio Based On Hardy-weinberg Equilibrium And Alternative Models Considering Sub-structuring And Relatedness. The Historical Development Of Low Template Mixture Analysis, Theory And Practice, Is Also Described, So The Reader Has A Full Understanding Of Rationale And Progression. Evaluation Of Evidence And Statement Writing Is Described In Detail, Along With Common Pitfalls And Their Avoidance. The Authors Have Been At The Forefront Of The Revolution, Having Made Substantial Contributions To Theory And Practice Over The Past Two Decades. All Methods Described Are Open-source And Freely Available, Supported By Sets Of Test-data And Links To Web-sites With Further Information. This Book Is Written Primarily For The Biologist With Little Or No Statistical Training. However, Sufficient Information Will Also Be Provided For The Experienced Statistician. Consequently, The Book Appeals To A Diverse Audience Covers Short Tandem Repeat (str) Analysis, Including Database Searching And Massive Parallel Sequencing (both Strs And Snps) Encourages Dissemination And Understanding Of Probabilistic Genotyping By Including Practical Examples Of Varying Complexity Written By Authors Intimately Involved With Software Development, Training At International Workshops And Reporting Cases Worldwide Using The Methods Described In This Book Front_cover Front-Matt_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of-Compl Copyrigh Content_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of-Complex- Contents About-the-aut_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of-Co About the authors Contributo_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of-Compl Contributors Forewor_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of-Complex- Foreword Prefac_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of-Complex-D Preface List-of-websites-and_2020_Forensic-Practitioner-s-Guide-to-the-Interpretatio List of websites and resources Chapter-1---Forensic-gene_2020_Forensic-Practitioner-s-Guide-to-the-Interpre 1 Forensic genetics: the basics 1.1 Short tandem repeat (STR) analysis 1.1.1 Historical development of multiplexed systems 1.2 Development and harmonization of European National DNA databases 1.2.1 Development of the European set of standard (ESS) markers 1.3 Hardy-Weinberg equilibrium 1.3.1 Measuring deviation from Hardy-Weinberg equilibrium Chi-square test Fisher's exact test 1.3.2 Extension to multiple alleles in STRs 1.4 Quality assurance of data 1.5 Recap: the laws of statistics 1.5.1 Effect of conditioning on probabilities 1.6 The likelihood ratio 1.7 Simple mixtures interpretation: the basics 1.7.1 Nomenclature 1.7.2 Minimum number of contributors 1.7.3 Interpretation of two-contributor mixtures 1.7.4 Step 1: condition the number of contributors 1.7.5 Step 2: state the alternative propositions 1.7.6 Step 3: evaluate the probability of the evidence under the defense proposition 1.7.7 Step 4: evaluate the probability of the crime-stain evidence under the prosecution proposition 1.7.8 Step 5: calculate the likelihood ratio 1.7.9 Step 6: report the likelihood ratio 1.8 Three allele mixtures 1.8.1 Calculation of the likelihood ratio 1.9 Two allele mixtures 1.10 Multiple contributors 1.11 Automating calculations with a computer program 1.11.1 Two-contributors 1.11.2 Recap 1.11.3 Extension to three or more contributors 1.12 The effect of population sub-structuring 1.12.1 What is a population? 1.12.1.1 Mutation 1.12.1.2 Migration 1.12.2 Genetic drift 1.12.3 Common ancestry 1.12.4 A computer simulation to illustrate effects of genetic drift 1.12.5 Extension to multi-allelic systems (STRs) 1.12.6 Balding and Nichols FST 1.12.6.1 Extension to mixtures 1.13 Simulation studies to show the effect of genetic drift and FST corrections 1.13.1 Details of the experiment 1.14 Sampling error correction 1.15 European Network of Forensic Science Institutes (ENFSI) evaluation of STR markers 1.15.1 Hardy-Weinberg tests 1.15.2 FST estimation 1.15.3 What FST to apply in casework? 1.16 Relatedness: kinship analysis 1.16.1 Effect of relatedness on the likelihood ratio calculation 1.16.2 Formal treatment and extension to mixtures 1.17 A brief introduction to the software approaches used to interpret mixtures 1.18 Summary Notes Chapter-2---Empirical-characte_2020_Forensic-Practitioner-s-Guide-to-the-Int 2 Empirical characterization of DNA profiles 2.1 Heterozygote balance 2.1.1 Peak height or peak area 2.1.2 Definitions 2.1.3 Methods to visualize data 2.1.4 The use of guidelines 2.1.5 Heterozygote balance summary 2.1.6 Characterization of heterozygote balance with STR-validator 2.2 Stutters 2.2.1 Characterization of stutters 2.2.2 Effect of number of STR repeats on the stutter size 2.2.3 Interpretation challenges with stutters 2.2.3.1 Filters applied by software 2.2.4 Stutters summary 2.2.5 Characterization of stutters with STR-validator 2.3 The Clayton guideline: a binary model to interpret simple mixtures 2.3.1 Step 1: identify the presence of a mixture 2.3.1.1 a) By the presence of extra alleles 2.3.1.2 b) Identify the presence of a mixture by peak imbalance 2.3.1.3 Genetic mutations affect the number of alleles visualized 2.3.1.4 Other genetic causes of heterozygote imbalance 2.3.1.5 Recognizing and dealing with rare genetic phenomena 2.3.2 Step 2: identify the potential number of contributors to a sample 2.3.3 Step 3: determine the approximate ratio of the components in the mixture 2.3.4 Step 4: determine the possible pairwise combinations for the components of the mixture 2.3.4.1 D13S317 2.3.4.2 D16S539 2.3.5 Step 5: compare with reference samples 2.4 Characterization of massively parallel sequencing (MPS) kits using STR-validator 2.5 Summary Chapter-3---Allele-drop-out-an_2020_Forensic-Practitioner-s-Guide-to-the-Int 3 Allele drop-out and the stochastic threshold 3.1 What causes allele drop-out? 3.1.1 A mathematical model showing the relationship of heterozygote balance and drop-out 3.2 Stochastic thresholds and drop-out 3.3 National DNA databases and low-template DNA crime stains 3.4 Definition of drop-out and extreme drop-out 3.5 Towards a meaningful definition of the low-template-DNA (T) threshold 3.6 Historical development of probabilistic evaluation of drop-out 3.7 Introducing probability of drop-out 3.7.1 The defense proposition 3.7.2 The prosecution proposition 3.8 A method to calculate probability of drop-out and the stochastic threshold 3.8.1 Experimental design and data collection 3.8.2 An explanation of logistic regression 3.9 Using STR-validator to plot drop-out characteristics 3.10 Summary Notes Chapter-4---Low-temp_2020_Forensic-Practitioner-s-Guide-to-the-Interpretatio 4 Low-template DNA 4.1 Historical 4.2 The consensus method explained 4.2.1 Contamination vs. drop-in 4.3 Statistical analysis 4.4 Working with replicates 4.4.1 An example to assess the strength of the evidence of three replicates 4.5 Degradation 4.5.1 Pre-PCR quantification and characterization of degradation 4.6 Summary Note Chapter-5---LRmix-mo_2020_Forensic-Practitioner-s-Guide-to-the-Interpretatio 5 LRmix model theory 5.1 Background 5.2 Model description 5.3 Theoretical considerations 5.4 Replicate probability 5.4.1 Drop-out and drop-in: definitions 5.4.2 Drop-out and drop-in: formalization 5.4.3 Example 1 5.4.4 Example 2 5.5 Genotype probability 5.5.1 Determining the genotypes for the unknowns 5.5.2 Correcting for population sub-structuring 5.5.2.1 Example 1 5.5.3 Example 3 5.5.4 Calculating genotype probabilities for relatedness 5.5.5 Example with relatedness in mixtures 5.6 Full example for calculating LR 5.7 Summary Notes Chapter-6---A-qualitative--semi-_2020_Forensic-Practitioner-s-Guide-to-the-I 6 A qualitative (semi-continuous) model: LRmix Studio 6.1 Interpretation of a major/minor DNA mixture, where the minor contributor is evidential 6.1.1 Case circumstances 6.1.2 Exploratory data analysis 6.1.3 The propositions 6.1.4 Sensitivity analysis 6.1.5 Calculation of levels of drop-out using an empirical method 6.1.6 Non-contributor analysis 6.1.7 How the LR is used in a statement 6.2 A case with two suspects 6.2.1 Case circumstances 6.2.2 Exploratory data analysis 6.2.3 Minimum number of contributors 6.2.4 Non-contributor tests 6.2.5 Formulation of propositions and exploratory data analysis 6.3 Considerations of the ISFG DNA commission 6.3.1 Formulation of propositions: summary 6.4 An example of a replicate analysis 6.5 Relatedness 6.6 Summary Notes Chapter-7---The-quantitative--_2020_Forensic-Practitioner-s-Guide-to-the-Int 7 The quantitative (continuous) model theory 7.1 Introduction 7.2 Towards a quantitative model 7.2.1 An example using the "MasterMix" Excel spreadsheet program 7.2.1.1 Step 1 7.2.1.2 Step 2 7.2.1.3 Step 3 7.2.2 Using the information across loci 7.3 Likelihood ratio estimation using probabilistic models 7.3.1 The normal distribution model 7.3.2 Step 1: calculation of expected peak heights 7.3.3 Step 2: calculate the weighting 7.3.4 Step 3: combine weightings with probability of genotype given the proposition 7.3.5 Estimating the parameters of the normal distribution: maximum likelihood estimation (MLE) 7.4 Recap 7.5 The Gamma model 7.5.1 Gamma distribution 7.5.2 Model for one contributor 7.5.3 Reparameterization 7.5.4 An example 7.5.4.1 A summary of parameter calculations 7.6 Drop-out for the gamma model explained 7.6.1 The Q allele 7.6.2 Comparison of the Excel spreadsheet with EuroForMix: a summary 7.7 Deconvolution 7.7.0.1 Calculation of marginal probabilities 7.8 Degradation 7.8.1 Demonstration with a spreadsheet 7.8.2 Recap 7.9 Stutter 7.9.1 Recap: taking account of peak height expectations using mixture proportions Mx 7.9.2 The full stutter model 7.9.3 A worked example showing the impact of stutter in EuroForMix 7.9.4 Combining stutter and degradation in the same model 7.9.5 Using a spreadsheet model to illustrate stutter 7.9.6 Dealing with forward and complex stutters 7.9.6.1 An example calculation using backward and forward stutter 7.10 The drop-in model 7.10.1 Estimating lambda using EuroForMix 7.10.2 Characterization of drop-in 7.11 Summary Notes Chapter-8---Euro_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of 8 EuroForMix 8.1 EuroForMix theory 8.2 Interpretation using EuroForMix 8.3 Theoretical considerations 8.4 Model features 8.5 An example 8.5.1 ENFSI exercise 1: a major/minor profile, where the minor component is evidential 8.5.1.1 Circumstances of the case 8.5.2 Reference samples 8.5.3 Preliminary assessment of the evidence 8.5.4 Model specification 8.5.5 The MLE fit tab 8.5.6 Model selection 8.5.7 Model validation 8.5.8 Probability-probability, PP-plots 8.5.9 LR sensitivity 8.5.10 Model fitted peak height tab 8.5.11 Non-contributor analysis 8.5.12 Deconvolution 8.6 ENFSI exercise 2: illustration of the advantage of using the quantitative model 8.6.1 Case circumstances 8.6.2 Effect of removing the conditioned suspect from the analysis 8.7 A complex case: robbery 8.7.1 Case circumstances 8.7.2 Analysis 8.7.3 EuroForMix analysis 8.8 Relatedness calculations 8.8.1 Results and discussion 8.9 Important updates to EuroForMix 8.10 A summary of the interpretation process Notes Chapter-9---Vali_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of 9 Validation 9.1 Context 9.2 Avoiding the black-box 9.3 Model validation 9.3.1 Conceptual validation 9.3.2 Operational validation 9.4 Defining benchmarks for LR-based models 9.4.1 Software validation 9.5 Validation study of LRmix Studio and EuroForMix 9.5.1 NGM DNA profiles 9.5.2 PPF6C DNA profiles 9.5.3 Replicates 9.5.4 Stutter filter 9.5.5 Allele frequency database 9.6 Design of experiments 9.6.1 Experiments using NGM profiles 9.6.2 Experiments using PPF6C profiles 9.7 Method 9.7.1 Defining characteristics of the models 9.8 Comparison of LRmix Studio and EuroForMix when the POI was the true contributor 9.9 Comparison of LRmix and EuroForMix when the POI was not a true contributor 9.10 Characterization of false positive results 9.10.1 NGM results 9.10.2 PPF6C results 9.11 Characterization of false negative results 9.12 LR values as a function of allele drop-out 9.12.1 NGM results 9.12.2 PPF6C results 9.13 Comparing the performance of two or more different models using ROC plots 9.14 Comparison of the stutter model in EuroForMix versus GeneMapper stutter filter 9.15 Calibration of the likelihood ratio 9.16 Further comparative studies between different software 9.17 Defining guidelines for using probabilistic genotyping software in forensic casework 9.18 Summary Notes Chapter-10---Development-and-_2020_Forensic-Practitioner-s-Guide-to-the-Inte 10 Development and implementation of DNAxs 10.1 DNA expert system 10.1.1 Functionalities of DNAxs 10.2 LoCIM-tool 10.3 DNAStatistX 10.3.1 Similarities and differences between EuroForMix and DNAStatistX 10.3.2 Optimizer to estimate parameters 10.3.3 Number of optimizer iterations 10.3.4 Overall versus dye-specific detection threshold 10.3.5 Model validation 10.3.6 DNAxs software testing 10.4 An exemplar case analyzed with DNAxs-DNAStatistX and EuroForMix 10.4.1 Analysis using DNAxs-DNAStatistX 10.4.2 Comparison of the exemplar using EuroForMix 10.5 Summary Chapter-11---Investigative-forensic-g_2020_Forensic-Practitioner-s-Guide-to- 11 Investigative forensic genetics: SmartRank, CaseSolver and DNAmatch2 11.1 National DNA databases 11.2 When is evaluative reporting appropriate? 11.3 A cautionary tale 11.4 Current methods used to compare crime stains with national DNA databases 11.4.1 Prüm inclusion rules 11.4.2 Prüm matching rules 11.4.3 CODIS inclusion and matching rules 11.5 Limitations of traditional national DNA databases: introducing "SmartRank" 11.6 Experimental details to test efficiency of SmartRank 11.6.1 SmartRank performance 11.7 SmartRank versus CODIS 11.7.1 Specifying a likelihood ratio threshold and a top ranked candidate list 11.7.2 Limitations 11.8 SmartRank exercise 11.9 Using EuroForMix for database searching 11.10 CaseSolver: an expert system based on EuroForMix 11.10.1 Method 11.10.1.1 Estimation of the number of contributors 11.10.2 Mixture comparison 11.10.2.1 Step 1: matching allele count 11.10.2.2 Step 2: qualitative LR approach 11.10.2.3 Step 3: quantitative LR approach 11.10.3 Model parameters 11.11 Demonstration using a real case example 11.11.1 Case circumstances 11.11.1.1 Crime stains and DNA profiles 11.11.2 Importing data to CS 11.11.3 Comparing references to mixtures 11.11.4 Viewing the results 11.11.5 Deconvolution 11.11.6 Relatedness searches 11.11.7 Advanced options 11.11.8 More detailed summary of results 11.11.8.1 Comparison against single-source profiles 11.11.8.2 Comparisons against mixtures 11.11.8.3 Relatedness 11.11.8.4 Potential false positives and negatives 11.11.9 Searching a large fictive national DNA database 11.12 The need for new digital tools in forensic genetics: improving efficiency of searching large DNA databases 11.12.1 Why use a stepwise strategy to extract candidate matches? 11.12.2 The flexibility of CS 11.12.3 Searching large national DNA databases 11.12.4 Using DNAmatch2 to search large databases and as a contamination search engine 11.12.5 Future improvements for estimating the number of contributors 11.13 Summary Notes Chapter-12---Interpretation--re_2020_Forensic-Practitioner-s-Guide-to-the-In 12 Interpretation, reporting, and communication 12.1 Early methods used to express the value of the evidence and their limitations 12.2 An outline of the important principles to interpret scientific evidence 12.3 Propositions 12.4 Scope of the forensic scientist as an investigator or evaluator 12.5 Hierarchy of propositions 12.6 Formulation of propositions to avoid bias 12.7 Activity level propositions 12.8 The prosecutor's fallacy and prior probabilities 12.8.1 Prior odds can be updated 12.9 Avoidance of the prosecutor's fallacy 12.10 Database searches 12.11 Statement writing 12.12 The use of the verbal scale and the "weak evidence effect" 12.13 Using terms like "inconclusive" and "excluded" as expressions in statements 12.14 Conditioning and the application to DNA mixtures 12.15 The effect of relatedness on the likelihood ratio 12.16 Assigning the number of contributors 12.17 Non-contributor tests 12.17.1 The "two suspect" problem 12.18 Summary Notes Chapter-13---Interpretation-of-complex-_2020_Forensic-Practitioner-s-Guide-t 13 Interpretation of complex DNA profiles generated by massively parallel sequencing 13.1 Introduction 13.2 SNP analysis 13.3 Number of contributors 13.4 Effect of choosing the wrong number of contributors 13.5 Comparison of quantitative and qualitative methods 13.6 Limitations 13.7 The effect of uncertainty of the number of contributors on the likelihood ratio 13.8 Summary 13.9 Short tandem repeats 13.9.1 Nomenclature 13.10 Historical perspective: lessons of the past 13.11 STR-MPS stutters 13.11.1 Extension to LUS+ nomenclature 13.12 Characterization of stutters with MPS 13.13 MPS programming using EuroForMix 13.13.1 Automated conversion of sequence into RU/LUS/LUS+ nomenclature 13.14 Demonstration of likelihood ratio calculations using EuroForMix 13.14.1 Analytical threshold (AT) 13.14.2 Rules for stutter filtering 13.14.3 Noise 13.14.4 Experimental summary 13.15 Automating the interpretation strategy 13.15.1 EuroForMix analysis 13.15.2 Stutter filter effect 13.16 Information gain: measuring the effect of the stutter filter and different nomenclatures 13.16.1 Information gain: comparing stutter filter vs no stutter filter 13.17 Summary Notes Appendix-A---Formal-descriptions_2020_Forensic-Practitioner-s-Guide-to-the-I A Formal descriptions of the genotype probabilities A.1 Extending the FST-formula to mixtures A.2 Relatedness A.2.1 Formulae for relatedness A.2.2 Extension with the FST-formula A.2.3 Specific relatedness formulae FST-correction Appendix-B---Formal-description_2020_Forensic-Practitioner-s-Guide-to-the-In B Formal description of the probabilistic models B.1 Definitions B.2 Extension to parameterized statistical models B.2.1 The likelihood for parameterized statistical models B.2.2 LR calculations for parameterized statistical models B.2.2.1 The maximum likelihood approach B.2.2.2 The Bayesian approach B.3 Mathematical details of the probabilistic models B.3.1 The contribution from assumed genotypes B.3.2 Mathematical details of the qualitative model B.3.3 Inference of the drop-out parameter B.3.3.1 A maximum likelihood approach B.3.3.2 A conservative approach B.3.4 Mathematical details of the quantitative model B.3.5 The gamma distribution B.3.6 Model for a single contributor B.3.6.1 Example 1 B.3.7 Model for multiple contributors B.3.7.1 Example 2 B.3.8 Model for allele drop-out B.3.8.1 Example 3 B.3.9 Model for allele drop-in B.3.9.1 Example 4 B.3.10 Model for multiple replicates B.3.11 Model for backward-stutters B.3.11.1 Example 5 B.3.12 Estimating the LR in EuroForMix B.3.12.1 A maximum likelihood approach B.3.12.2 An integration-based approach ("Full Bayesian") B.3.12.3 A conservative approach B.4 Deconvolution Bibliograp_2020_Forensic-Practitioner-s-Guide-to-the-Interpretation-of-Compl Bibliography Index Index Back_cover "Over the past twenty years, there’s been a gradual shift in the way forensic scientists approach the evaluation of DNA profiling evidence that is taken to court. Many laboratories are now adopting ‘probabilistic genotyping’ to interpret complex DNA mixtures. However, current practice is very diverse, where a whole range of technologies are used to interpret DNA profiles and the software approaches advocated are commonly used throughout the world. Forensic Practitioner’s Guide to the Interpretation of Complex DNA Profiles places the main concepts of DNA profiling into context and fills a niche that is unoccupied in current literature. The book begins with an introduction to basic forensic genetics, covering a brief historical description of the development and harmonization of STR markers and national DNA databases. The laws of statistics are described, along with the likelihood ratio based on Hardy-Weinberg equilibrium and alternative models considering sub-structuring and relatedness. The historical development of low template mixture analysis, theory and practice, is also described, so the reader has a full understanding of rationale and progression. Evaluation of evidence and statement writing is described in detail, along with common pitfalls and their avoidance. The authors have been at the forefront of the revolution, having made substantial contributions to theory and practice over the past two decades. All methods described are open-source and freely available, supported by sets of test-data and links to web-sites with further information. This book is written primarily for the biologist with little or no statistical training. However, sufficient information will also be provided for the experienced statistician. Consequently, the book appeals to a diverse audience."-- Provided by publisher