Presentation on theme: "Forensic DNA profiling workshop"— Presentation transcript:
1 Forensic DNA profiling workshop Dan E. Krane, Wright State University, Dayton, OhioJason R. Gilder, Forensic Bioinformatics, Fairborn, OhioForensic Bioinformatics (
2 I: Overview of what DNA tests can do for: A. Prosecution B. Defense C I: Overview of what DNA tests can do for: A. Prosecution B. Defense C. Post-conviction testing
3 DNA Technology in Court Criminal ProsecutionUnprecedented sensitivity and specificity for typing biological samplesGrowing use of databanks and dragnets to identify suspectsRapidly becoming cheaper and faster
4 Possible DNA SourcesSlide to show that with STR’s we can do “non-conventional” evidence such as dandruff and skin cells from the handle of a knife.
5 DNA Technology in Court Criminal DefenseUnprecedented sensitivity and specificity for typing biological samplesPotential support for alternative theories of the case
6 DNA Technology in Court Post-conviction exonerations (208 in US) based on DNA evidence have revealed problems with the justice system
8 Three generations of DNA testing RFLPAUTORADAllele = BANDDQ-alphaTEST STRIPAllele = BLUE DOTAutomated STRELECTROPHEROGRAMAllele = PEAK
9 Two relatively new DNA tests Mitochondrial DNAmtDNA sequenceSensitive but not discriminatingY-STRsUseful with mixturesPaternally inherited
10 Phenotyping: DNA WitnessTM New test by DNA Print Genomics (Florida)Tests SNPs (single nucleotide polymorphisms)Identifies ‘genetic heritage’ of sampleProvides percentages of makeup:Sub-Saharan AfricanEast-AsianIndo-EuropeanNative-AmericanLatest versions infer hair and eye colorHas been used in assisting investigations
11 Lab-on-a-chip Currently in R&D Use microdevices to Extract DNA Quantify DNAPCR amplify DNACapillary electrophoresisAll on small, integrated glass or plastic chipQuick (test in half hour?)Very small samples?Portable?
12 Basic terminology: Genetics DNA Polymorphism (“many forms”)Regions of DNA which differ from person to personLocus (plural = loci)Site or location on a chromosomeAlleleDifferent variants which can exist at a locusDNA ProfileThe combination of alleles for an individual
13 Basic terminology: Technology Amplification or PCR (Polymerase Chain Reaction)A technique for ‘replicating’ DNA in the laboratory (‘molecular Xeroxing’)Region to be amplified defined by PRIMERSCan be ‘color coded’ElectrophoresisA technique for separating molecules according to their size
15 STR Short tandem repeat Describes a type of DNA polymorphism in which: a DNA sequence repeatsover and over againand has a short (usually 4 base pair) repeat unitA length polymorphism -- alleles differ in their length3 repeats: AATG AATG AATG4 repeats: AATG AATG AATG AATG5 repeats: AATG AATG AATG AATG AATG6 repeats: AATG AATG AATG AATG AATG AATG
16 Crime Scene Samples & Reference Samples Extract and purify DNADifferential extraction in sex assault cases separates out DNA from sperm cells
17 Extract and Purify DNA Warm soapy water Releases biological material Organic ExtractionEthanol Precipitation
18 PCR AmplificationGroups of amplified STR products are labeled with different colored dyes (blue, green, yellow)
20 ABI 310 Genetic Analyzer: Capillary Electrophoresis Amplified STR DNA injected onto columnElectric current appliedDNA pulled towards the positive electrodeDNA separated out by size:Large STRs travel slowerSmall STRs travel fasterDetectorWindowColor of STR detected and recorded as it passes the detector
22 Reading an electropherogram ALLELE CALLSLoci arranged by size and colorAllele designation given as a numberD3vWAFGAD8D21D18D5D13D7NUMBER OF PEAKS1 peak = homozygous2 peaks = heterozygous3 or more peaks = mixed sample (?)AmelogeninHEIGHT OF PEAKProportional to amount of allele (approx)RFU (relative fluorescent units)AmelogeninSex of sampleXY = MaleX = female
27 Statistical estimates: the product rule 0.222x0.222x2= 0.1
28 Statistical estimates: the product rule 1 in 101 in 22,200x1 in 1111 in 20= 0.11 in 1001 in 141 in 811 in 113,400x1 in 1161 in 171 in 161 in 31,552x1 in 79,531,528,960,000,0001 in 80 quadrillion
29 What more is there to say after you have said: “The chance of a coincidental match is one in 80 quadrillion?”
30 What more is there to say after you have said: “The chance of a coincidental match is one in 80 quadrillion?”• Two samples really do have the same sourceSamples match coincidentallyAn error has occurred
31 DNA match probability Random Match Probability (RMP) What is the chance of finding a random, unrelated person in a given population that has a given DNA profile?NOT the probability that the defendant is guiltyNOT the probability that someone other than the defendant committed the crime
32 State of Texas v. Josiah Sutton (1999) Woman raped in car by two menIdentifies Sutton and AdamsDNA test on vaginal swab, pubic combings, jeans, and semen stain on car seat
34 NRC I report"To say that two patterns match, without providing any scientifically valid estimate (or, at least, an upper bound) of the frequency with which such matches might occur by chance, is meaningless.”National Research Council. DNA Technology in Forensic Science pg. 9
35 Sample DQA LDLR GYPA HBGG D7 GC D1S80 1.1, 2, 3, 4.1 (1.2) AB AC B VSF1.1, 2, 3, 4.1 (1.2)ABACB20,21,24,25,28Sutton1.1, 2A25,28Complainant3,4.2/4.321,28FrequencyOf All Included Genotypes0.681.000.580.530.33Cumulative Probability of Inclusion0.0691 in 15
36 RetestsSperm fraction of vaginal sample shows profiles of two male donorsSutton excluded at seven of nine lociSutton exoneratedAdditional tests confirm profiles of two unknown men on jeans
37 Difference between how suspect was Identified Confirmatory ID CaseSuspect is first identified by non-DNA evidenceDNA evidence is used to corroborate traditional police investigationCold Hit CaseSuspect is first identified by search of DNA databaseDNA evidence is used to identify suspect as perpetrator, to exclusion of others, from the outsetTraditional police work is no longer focus
38 The Problem: Ascertainment bias First three approaches differ in how they take into account ascertainment bias, a byproduct of identifying an individual from a database search.Ascertainment bias is statistical effect of fact suspect first identified by search of a databaseHow must RMP be modifiedNew Piece of dataHow to calculate chance observation is coincidental
39 NRC I & NRC IIPosition: Both say ascertainment bias makes the link between suspect and crime scene DNA weaker—less probative.Rationale: As the size of the database searched increases, so does the chance that you will find a match to the crime scene profile by chance.
40 NRC I & NRC IIExample: If you are looking for someone named “Rembrandt,” the likelihood of finding match(es) greatly increases if you search Dutch census data versus a local phone book. And your amazement at finding another “Rembrandt” decreases as database increases
41 Cold Hit StatisticsNRC I—test additional loci and report F for those loci onlyPresumes ascertainment bias is a serious problemNRC II—report FxN, where N is the number of profiles in the databasee.g., if F=1 in 1 billion; N=1 million; then tell jury RMP=1 in 1000Friedman, Balding, Donnelly, Weir (and prosecutors everywhere)—ascertainment bias is not a problem, so just tell the jury F
42 Balding/Donnelly Position A DNA database search is not a multiple-opportunity search, it is a multitude of single-opportunity searchesAlthough there are multiple opportunities to match someone, there is only a single opportunity to match your client, therefore RMP=F for the defendantIs this position generally accepted?What is the relevant question?
43 Problems with Balding/Donnelly Position Some database searches do create multiple opportunities to incriminate the same persone.g., suspect’s profile searched against multiple items of evidence from multiple unsolved crimesB/D assume probability of guilt in a cold hit case may be low, notwithstanding tiny value of F, because prior probability is lowWill jurors understand (and share) this assumption?Failure to consider probability of error
44 The False Positive Fallacy “If the probability of a false positive is one in a thousand that means there are 999 chances in 1000 we have the right guy.”Not necessarily true; probability of “having right guy” depends on strength of all the evidenceIf prior odds of guilt are 1:1000 and odds of a false positive are 1:1000, then chances of “having the right guy” are 50:50 (even odds)See, Thompson, Taroni & Aitken, JFS, 2003.
45 Inadvertent Transfer of DNA Primary transfer -- from individual to an object or another personR. van Oorschot & M. Jones, DNA fingerprints from fingerprints. Nature, 387: 767 (1997).Secondary transfer -- from the point of primary transfer to a second object or person“…in some cases, material from which DNA can be retrieved is transferred from object to hand.” Id.
46 Quantities of DNA Optimum amount of template: 0.5 to 2.0 ng 6 to 7 pg of DNA in each diploid human cellOur bodies are made of many billions if not trillions of cellspg = picogram (milligram, microgram, nanogram, picogram)SGM+ and Profiler Plus test kits are designed to fail with less than 100 pg to minimize these problems
47 DNA content of biological samples: Type of sampleAmount of DNABlood30,000 ng/mLstain 1 cm in area2200 ngstain 1 mm in area22 ngSemen250,000 ng/mLPostcoital vaginal swab0 - 3,000 ngHairpluckedng/hairshedng/hairSaliva5,000 ng/mLUrineng/mL
48 Taylor & Johnson Studies (1) A kisses B on cheekC touches B’s cheek with a gloveDNA consistent with A and B found on glove
49 Taylor & Johnson Studies (2) A wipes his own face with a damp towelB wipes her face with same towelC touches B’s face with gloveDNA consistent with A and B found on glove
50 Documenting errors: DNA Advisory Board Quality Assurance Standards for Forensic DNA Testing Laboratories, Standard 14[Forensic DNA laboratories must] “follow procedures for corrective action whenever proficiency testing discrepancies and/or casework errors are detected” [and] “shall maintain documentation for the corrective action.”
58 Documenting errors Suspect doesn’t match himself . . . . but then, staff is “‘always’ getting people’s names wrong”:
59 The science of DNA profiling is sound The science of DNA profiling is sound. But, not all of DNA profiling is science.
60 Opportunities for subjective interpretation? Can “Tom” be excluded?Suspect D3 vWA FGATom 17, 17 15, 17 25, 2560
61 Opportunities for subjective interpretation? Can “Tom” be excluded?Suspect D3 vWA FGATom 17, 17 15, 17 25, 25No -- the additional alleles at D3 and FGA are “technical artifacts.”61
62 Opportunities for subjective interpretation? Can “Dick” be excluded?Suspect D3 vWA FGATom 17, 17 15, 17 25, 25Dick 12, 17 15, 17 20, 2562
63 Opportunities for subjective interpretation? Can “Dick” be excluded?Suspect D3 vWA FGATom 17, 17 15, 17 25, 25Dick 12, 17 15, 17 20, 25No -- stochastic effects explain peak height disparity in D3; blob in FGA masks 20 allele.63
64 Opportunities for subjective interpretation? Can “Harry” be excluded?Suspect D3 vWA FGATom 17, 17 15, 17 25, 25Dick 12, 17 15, 17 20, 25Harry 14, 17 15, 17 20, 25No -- the 14 allele at D3 may be missing due to “allelic drop out”; FGA blob masks the 20 allele.64
65 Opportunities for subjective interpretation? Can “Sally” be excluded?Suspect D3 vWA FGATom 17, 17 15, 17 25, 25Dick 12, 17 15, 17 20, 25Harry 14, 17 15, 17 20, 25Sally 12, 17 15, 15 20, 22No -- there must be a second contributor; degradation explains the “missing” FGA allele.65
66 Subjective interpretation and statistics Frequency estimates (for Tom):p x 2pq x p2Suspect D3 vWA FGATom 17, 17 15, 17 25, 25Dick 12, 17 15, 17 20, 25Harry 14, 17 15, 17 20, 25Sally 12, 17 15, 15 20, 2266
67 The science of DNA profiling is sound The science of DNA profiling is sound. But, not all of DNA profiling is science.67
68 Sources of ambiguity in STR interpretation DegradationAllelic dropoutFalse peaksMixturesAccounting for relativesThreshold issues -- marginal samples
69 Degradation SMALL LARGE When biological samples are exposed to adverse environmental conditions, they can become degradedWarm, moist, sunlight, timeDegradation breaks the DNA at randomLarger amplified regions are affected firstClassic ‘ski-slope’ electropherogramPeaks on the right lower than peaks on the left
70 Sources of ambiguity in STR interpretation DegradationAllelic dropoutFalse peaksMixturesAccounting for relativesThreshold issues -- marginal samples
71 ? Allelic Dropout 1500 150 Reference sample Evidence sample Peaks in evidence samples all very lowMostly below 150 rfuPeaks in reference sample much higherAll well above 800 rfuAt D13S817:Reference sample: 8, 14Evidence sample: 8, 814 allele has dropped out -- or has it?Tend to see with ‘marginal samples’
72 Sources of ambiguity in STR interpretation DegradationAllelic dropoutFalse peaksMixturesAccounting for relativesThreshold issues -- marginal samples
73 Not all signal comes from DNA associated with an evidence sample • Stutter peaksPull-up (bleed through)Spikes and blobs
79 Interpreting mixtures Four alleles – two or more peoplePeak height balance makes it more likely that there are three or more contributorsDifficult to determine number of contributors
80 Using given references Cheung 16Wong 15,17Whitney 14,16
81 Other possibilities Tom 15,16 Dick 16,17 Harry 14,16 Cheung 16 Wong 15,17Whitney 14,16
82 Possibly four contributors Joe 16,17Bob 15,16Barry 16Fred 14Tom 15,16Dick 16,17Mary 14,16Cheung 16Wong 15,17Whitney 14,16
83 Determining the number of contributors Mixtures can exhibit up to two peaks per contributor at any given locusMixtures can exhibit as few as 1 peak at any given locus (regardless of the number of contributors)
85 Allele counting method Many labs determine the number of contributors to a mixed sample based on maximum allele counts at a locus3-4 peaks = 2 contributors5-6 peaks = 3 contributors7-8 peaks = 4 contributors
86 How many contributors to a mixture? How many contributors to a mixture if analysts can discard a locus?Maximum # of alleles observed in a 3-person mixture# of occurrencesPercent of cases20.0037844,967,0343.39593,037,01063.49648,532,03733.123,3987,274,823112,469,39826,788,5400.004.9676.7518.28There are 45,139,896 possibledifferent 3-way mixtures of the 648individuals in the MN BCI database.There are 146,536,159 possible different 3-person mixtures of the 959 individuals in the FB I database (Paoletti et al., November 2005 JFS).
87 How many contributors to a mixture? Maximum # of alleles observed in a 4-person mixture# of occurrencesPercent of cases413,4800.0258,596,32015.03635,068,04061.30712,637,10122.098896,4351.57There are 45,139,896 possibledifferent 3-way mixtures of the 648individuals in the MN BCI database.There are 57,211,376 possible different 4-way mixtures of the 194 individuals in the FBI Caucasian database (Paoletti et al., November 2005 JFS). (35,022,142,001 4-person mixtures with 959 individuals.)
88 Sources of ambiguity in STR interpretation DegradationAllelic dropoutFalse peaksMixturesAccounting for relativesThreshold issues -- marginal samples
89 What contributes to overlapping alleles between individuals? • Identity by state-- many loci have a small number of detectable alleles (only 6 for TPOX and 7 for D13, D5, D3 and TH01)-- some alleles at some loci are relatively common• Identity by descent-- relatives are more likely to share alleles than unrelated individuals-- perfect 13 locus matches between siblings occur at an average rate of 3.0 per 459,361 sibling pairs
91 Allele sharing in databases • Original FBI dataset’s mischaracterization rate for 3-person mixtures (3.39%) is more than two above the average observed in five sets of randomized individuals• Original FBI dataset has more shared allele counts above 19 than five sets of randomized individuals (3 vs. an average of 1.4)
92 Sources of ambiguity in STR interpretation DegradationAllelic dropoutFalse peaksMixturesAccounting for relativesThreshold issues -- marginal samples
93 Where do peak height thresholds come from (originally)? • Applied Biosystems validation study of 1998• Wallin et al., 1998, “TWGDAM validation of the AmpFISTR blue PCR Amplification kit for forensic casework analysis.” JFS 43:
94 Where do peak height thresholds come from (originally)?
95 Where do peak height thresholds come from? • “Conservative” thresholds established during validation studies• Eliminate noise (even at the cost of eliminating signal)• Can arbitrarily remove legitimate signal• Contributions to noise vary over time (e.g. polymer and capillary age/condition)Analytical chemists use LOD and LOQ
96 Signal Measure Saturation Measured signal (In Volts/RFUS/etc) Quantification limitμb + 10σbμb + 3σbDetection limitMean backgroundSignalμb
98 RFU levels at all non-masked data collection points Taken from a single negative control (one of the 50 used in this validation study). This distribution is from a blue channel and exhibits an average baseline of 5.5 RFUs. Each negative control has an average of 5,932 data collection points (after masking) with a standard deviation of 131 per run.
99 LOD/LOQ validation study Average (μb) and standard deviation (σb) values with corresponding LODs and LOQs from positive, negative and reagent blank controls in 50 different runs. BatchExtract: ftp.ncbi.nlm.nih.gov/pub/forensics/ LOD/LOQ software:
101 Lines in the sand: a two-person mix? Two reference samples in a 1:10 ratio (male:female). Three different thresholds are shown: 150 RFU (red); LOQ at 77 RFU (blue); and LOD at 29 RFU (green).
102 Observer effects, aka expectation effects --the tendency to interpret data in a manner consistent with expectations or prior theories (sometimes called “examiner bias”)Most influential when:Data being evaluated are ambiguous or subject to alternate interpretationsAnalyst is motivated to find a particular result
106 Analyst often have strong expectations about the data DNA Lab Notes (Commonwealth v. Davis)“I asked how they got their suspect. He is a convicted rapist and the MO matches the former rape…The suspect was recently released from prison and works in the same building as the victim…She was afraid of him. Also his demeanor was suspicious when they brought him in for questioning…He also fits the general description of the man witnesses saw leaving the area on the night they think she died…So, I said, you basically have nothing to connect him directly with the murder (unless we find his DNA). He said yes.”
107 Analyst often have strong expectations about the data DNA Lab Notes“Suspect-known crip gang member--keeps ‘skating’ on charges-never serves time. This robbery he gets hit in head with bar stool--left blood trail. Miller [deputy DA] wants to connect this guy to scene w/DNA …”“Death penalty case! Need to eliminate Item #57 [name of individual] as a possible suspect”
108 Analysts’ expectations may lead them to: Resolve ambiguous data in a manner consistent with expectationsMiss or disregard evidence of problemsMiss or disregard alternative interpretations of the dataThereby undermining the scientific validity of conclusionsSee, Risinger, Saks, Thompson, & Rosenthal, The Daubert/Kumho Implications of Observer Effects in Forensic Science: Hidden Problems of Expectation and Suggestion. 93 California Law Review 1 (2002).
109 What is LCN?DNA profiling performed at or beneath the stochastic thresholdTypically less than 0.5 ng of DNA templateTypically involves modifications of the testing methodology (e.g. increased polymerase; additional rounds of amplification; skipping quantitation)Consensus profiles
110 Applied Biosystems SGM Plus User’s Manual p.1-14
111 “The PCR amplification parameters have been optimized to produce similar peak heights within and between loci. The peak height generated at a locus for a heterozygous individual should be similar between the two alleles. The kit is also designed to generate similar peak heights between loci labeled with the same dye so that each locus will have approximately the same sensitivity.”Applied Biosystems SGM Plus User’s Manual p.1-13
112 What is LCN?DNA profiling performed at or beneath the stochastic thresholdTypically less than 0.5 ng of DNA templateTypically involves modifications of the testing methodology (e.g. increased polymerase; additional rounds of amplification; skipping quantitation)Consensus profiles
113 Ways of increasing sensitivity Increasing rounds of PCRMore than 28 (typically 32 or 34)Increasing injection timeExtra purification stepsCellmark Montage,LGC Forensics EnhancementAdd more amplified product
114 Applied Biosystems SGM Plus User’s Manual p.1-14
115 “The PCR amplification parameters have been optimized to produce similar peak heights within and between loci. The peak height generated at a locus for a heterozygous individual should be similar between the two alleles. The kit is also designed to generate similar peak heights between loci labeled with the same dye so that each locus will have approximately the same sensitivity.”Applied Biosystems SGM Plus User’s Manual p.1-13
116 Stochastic effectsUltimately due to poor statistical sampling of underlying templateThe four horsemen of stochasticismExaggerated stutterExaggerated peak height imbalance (0 to 100%)Allelic drop-out (extreme peak height imbalance)Allelic drop-in (contamination)
118 Stochastic effectsUltimately due to poor statistical sampling of underlying templateThe four horsemen of stochasticismExaggerated stutter (up to 50%)Exaggerated peak height imbalance (0 to 100%)Allelic drop-out (extreme peak height imbalance)Allelic drop-in (contamination)
121 How helpful is quantitation? Optimum amount of template: 0.5 to 2.0 ng6 to 7 pg of DNA in each diploid human cellIn a mixed sample containing 0.5 ng of template, less than 0.5 ng comes from each contributor
124 Consensus profilesAlleles are not reported unless they are seen in at least two runsConsidering two runs serves as a safeguard against allelic drop-in (contamination)Considering three or more runs begins to safeguard against drop-outIf a sample is being split four or more times, shouldn’t conventional tests be done?
125 Replicate analysis runs D3vWAD16D2D8D21D18D19THO1FGAAmp 116s 1717 F10 13(17) 20(10) 1330 F(12) (17)14 (15) 16(7) 9.321 24Amp 215s 16 1712 (13)(9) 10 (11) 1315 169.3Consensus16 1713 FF F14 159.3 F21, 24An allele is considered to be legitimate if it is observed twice.
126 Sources of ambiguity in STR interpretation DegradationAllelic dropoutFalse peaksMixturesAccounting for relativesThreshold issues -- marginal samples
130 Familial searchingDatabase search yields a close but imperfect DNA matchCan suggest a relative is the true perpetratorGreat Britain performs them routinelyReluctance to perform them in US since 1992 NRC reportCurrent CODIS software cannot perform effective searches
131 Three approaches to familial searches Search for rare alleles (inefficient)Count matching alleles (arbitrary)Likelihood ratios with kinship analyses
133 Is the true DNA match a relative or a random individual? Given a closely matching profile, who is more likely to match, a relative or a randomly chosen, unrelated individual?Use a likelihood ratio
134 Probabilities of siblings matching at 0, 1 or 2 alleles HF = 1 for homozygous loci and 2 for heterozygous loci; Pa is the frequency of the allele shared by the evidence sample and the individual in a database.
135 Probabilities of parent/child matching at 0, 1 or 2 alleles HF = 1 for homozygous loci and 2 for heterozygous loci; Pa is the frequency of the allele shared by the evidence sample and the individual in a database.
136 Other familial relationships Cousins:Grandparent-grandchild; aunt/uncle-nephew-neice;half-sibings:HF = 1 for homozygous loci and 2 for heterozygous loci; Pa is the frequency of the allele shared by the evidence sample and the individual in a database.
137 Is the true DNA match a relative or a random individual? This more difficult question is ultimately governed by two considerations:What is the size of the alternative suspect pool?What is an acceptable rate of false positives?
138 III: What can go wrong and where problems might occur
139 Victorian Coroner’s inquest into the death of Jaidyn Leskie Toddler disappears in bizarre circumstances: found dead six months laterMother’s boy friend is tried and acquitted.Unknown female profile on clothing.Cold hit to a rape victim.RMP: 1 in 227 million.Lab claims “adventitious match.”
140 Victorian Coroner’s inquest into the death of Jaidyn Leskie Condom with rape victim’s DNA was processed in the same lab 1 or 2 days prior to Leskie samples.Additional tests find matches at 5 to 7 more loci.Review of electronic data reveals low level contributions at even more loci.Degradation study further suggests contamination.
141 Degradation, inhibition SMALLLARGEWhen biological samples are exposed to adverse environmental conditions, they can become degradedWarm, moist, sunlight, timeDegradation breaks the DNA at randomLarger amplified regions are affected firstClassic ‘ski-slope’ electropherogramDegradation and inhibition are unusual and noteworthy.
142 Degradation, inhibition The Leskie Inquest, a practical applicationUndegraded samples can have “ski-slopes” too.How negative does a slope have to be to an indication of degradation?Experience, training and expertise.Positive controls should not be degraded.
143 Degradation, inhibition The Leskie InquestDNA profiles in a rape and a murder investigation match.Everyone agrees that the murder samples are degraded.If the rape sample is degraded, it could have contaminated the murder samples.Is the rape sample degraded?
145 Victorian Coroner’s inquest into the death of Jaidyn Leskie “8. During the conduct of the preliminary investigation (before it was decided to undertake an inquest) the female DNA allegedly taken from the bib that was discovered with the body was matched with a DNA profile in the Victorian Police Forensic Science database. This profile was from a rape victim who was subsequently found to be unrelated to the Leskie case.”
146 Victorian Coroner’s inquest into the death of Jaidyn Leskie “8. The match to the bib occurred as a result of contamination in the laboratory and was not an adventitious match. The samples from the two cases were examined by the same scientist within a close time frame.”
147 The science of DNA profiling is sound The science of DNA profiling is sound. But, not all of DNA profiling is science. This is especially true in situations involving: small amounts of starting material, mixtures, relatives, and analyst judgment calls.
148 Steps in Preparing a DNA Case Obtain all lab reportsRed flags:unfamiliar techniquesequivocal matches (profile “similar but cannot be definitively included nor excluded”);contingent matches (profile included “if…” or “but…”;partial/incomplete profiles;mixtures;unusually modest statistics; no statistics; likelihood ratios
149 Steps in Preparing a DNA Case Initial discoveryFull history of all samples from collection to current dispositionComplete DNA lab notes (bench notes)Electronic dataAnalysts’ credentials, proficiency test recordLab’s incidence reports; unexpected event files; accreditation filesObtain expert assistance for initial review
150 Steps in Preparing a DNA Case Initial evaluation of caseIdentify possible lines of attackAdditional/alternative experts neededNeeds for follow-up discovery—e.g., validation; proficiency problems; error problemsConsider advisability of additional testingReplications; untested items; other experimentsFinal evaluation of strategyConsider ways to blunt/deflect prosecution (or defense) testimonyPrepare exhibits, lines of examination, motions in limine; notices of objection, etc.
151 ResourcesInternetForensic Bioinformatics Website:Applied Biosystems Website: (see human identity and forensics)STR base: (very useful)Books‘Forensic DNA Typing’ by John M. Butler (Academic Press)ScientistsLarry Mueller (UC Irvine)Simon Ford (Lexigen, Inc. San Francisco, CA)William Shields (SUNY, Syracuse, NY)Mike Raymer and Travis Doom (Wright State, Dayton, OH)Marc Taylor (Technical Associates, Ventura, CA)Keith Inman (Forensic Analytical, Haywood, CA)Testing laboratoriesTechnical Associates (Ventura, CA)Forensic Analytical (Hayward, CA)Other resourcesForensic Bioinformatics (Dayton, OH)