Brief History of Forensic DNA Typing

Brief History of Forensic DNA Typing
1980 - Ray White describes first polymorphic RFLP marker
1985 - Alec Jeffreys discovers multilocus VNTR probes
1985 - first paper on PCR
1988 - FBI starts DNA casework
1991 - first STR paper
1995 - FSS starts UK DNA database
1998 - FBI launches CODIS database
DNA Use in Forensic Cases
Most are rape cases (>2 out of 3)
Looking for match between evidence and suspect
Must compare victim’s DNA profile

Mixtures must be resolved
DNA is often degraded
Inhibitors to PCR are often present
Challenges
Why DNA?
Individually unique (except for identical twins)

Present in virtually all body tissues (bone, hair, blood, saliva and other body fluids and tissues)

Inheritance (Paternity, Maternity, Unidentified Remains)

Crime incidents and scenes (exchange principle)
Human Identity Testing
Forensic cases -- matching suspect with evidence
Paternity testing -- identifying father
Historical investigations
Missing persons investigations
Mass disasters -- putting pieces back together
Military DNA “dog tag”
Convicted felon DNA databases
Sample Obtained from Crime Scene or Paternity Investigation
Biology
Technology
Genetics
Steps in DNA Sample Processing
Sources of Biological Evidence
Blood
Semen
Saliva
Urine
Hair
Teeth
Bone
Tissue
DNA in the Cell
DNA Amplification with the Polymerase Chain Reaction (PCR)
In 32 cycles at 100% efficiency, 1.07 billion copies of targeted DNA region are created
PCR Copies DNA Exponentially through Multiple Thermal Cycles
Short Tandem Repeats (STRs)
the repeat region is variable between samples while the flanking regions where PCR primers bind are constant
7 repeats
8 repeats
Homozygote = both alleles are the same length
Heterozygote = alleles differ and can be resolved from one another
D2s44 locus  - a highly variable locus
 
There are 28 different alleles of the D2s44 locus
 
so for two people to have the identical genotype if all the alleles are equally common:
1/28 = .03
if you test 10 genes, each with 28 alleles…
 
..03 x .03 x .03….. (.03)10 = 6 x 10-16 a very small
chance of a false positive
Review of STR DNA Fingerprinting Process;
Probability
Underlying Principles for Forensic Probabilities
Hardy-Weinberg proportions
assuming random mating and excluding population substructure
homozygote frequencies described by p2
heterozygote frequencies described by 2pq
Linkage Equilibrium
assuming random mating and sufficient genetic distance
allows frequencies of individual loci to be multiplied to derive the frequency of the profile as a whole
DNA Database
Locus A Locus B
Allele Freq (%) Allele Freq (%)
13 25 8 20
14 15 9 25
15 20 10 10
16 1 11 15
17 39 12 5
13 25
With the previous data set, determine:

The estimated frequency of homozygotes of the 16 allele from locus A.
homozygotes = p2 = (.01)2 = .0001= 1 / 10,000


The estimated frequency of heterozygotes of the 10 and 12 alleles from locus B.
heterozygotes = 2pq = 2(.1)(.05) = .01 =1/100
With the previous data set, determine:
The frequency of the profile:
Locus A - 16,16
Locus B - 10,12

from previous slide
Locus A - 16,16 = .0001
Locus B - 10,12 = .01

(.0001)(.01) = .000001 = 1/1,000,000
In words…
On average, the above profile is expected to be observed once per one million profiles tested.

In court...
these numbers are typically compared to the world population (~6 billion) at which point their significance becomes immediately apparent.

Different primer sets produce different PCR product sizes for the same STR allele
TCAT repeat unit
Analysis of ALU polymorphism
Forensic sample’s (E) genotype - homozygous -/-
Exclusion of 2 homozygous +/+subjects (contain ALU sequence)
Inclusion of 6 homozygous -/- subjects (does not contain ALU sequence)
Picture of cheek ALU gel
Picture of hair ALU gel
1 2 3 4 5 6 7 8 9 A B C D E
E D C B A 9 8 4 3 2 1
5
6
7
ALU polymorphism comparison table
U.S. vs. our sample
ALU polymorphism comparison


Sampling of US citizens ALU frequencies
Our study of ALU frequencies
Defies Hardy-Weinberg equilibrium theory (p2 +2pq+q2=1.00)
Non-random sample
Analysis of VNTR Polymorphism
Picture of VNTR from hair samples.
Picture of VNTR from cheek samples.
Exclusion of four samples for improper orientation with the forensic sample .
Two potential suspects remain (P1 is real match)
Non-conclusive results

1 2 3 4 5 6 7 8 9 10
1 2 3 4 5 6 7 8 9 10
Conclusion
No identification for mystery sample due to non-conclusive results of two potential matches.
Eliminated six out of eight samples
Unreliability in PCR and DNA Typing

Suggestions
Extra DNA Typing trials
Less human error: incorrect pipette use, loading gels, cross-contamination


separation
matrix (based on size)
allele 1 size n(4R)
allele 2 size n(3R)
allele 3 size n(2R)
allele 4 size n(1R)
n(1R) n(2R) n(3R) n(4R)
Large
Small
allele 1 size n(4R)
Large
Small
allele 3 size n(2R)
isolate DNA
amplify (PCR)
separate and detect
You can obtain DNA from any source
DNA is quite stable and can be obtained relatively intact from some specimens that are thousands of years old
Multiplex PCR
Over 10 Markers Can Be Copied at Once
Sensitivities to levels less than 1 ng of DNA
Ability to Handle Mixtures and Degraded Samples
Different Fluorescent Dyes Used to Distinguish STR Alleles with Overlapping Size Ranges

An Example Forensic STR Multiplex Kit
Of the 13 loci that can be used for forensics work the most common profile has a frequency of occurrence of 1 in 250 billion; the rarest has a frequency of occurrence of 1 in 7 x1039
Available Kits for STR Analysis
Kits make it easy for labs to just add DNA samples to a pre-made mix
13 CODIS core loci
Profiler Plus and COfiler (PE Applied Biosystems)
PowerPlex 1.1 and 2.1 (Promega Corporation)
Increased power of discrimination
CTT (1994): 1 in 410
SGM Plus™ (1999): 1 in 3 trillion
PowerPlex ™ 16 (2000): 1 in 2 x 1017
ABI Prism 310 Genetic Analyzer
Close-up of ABI Prism 310 Sample Loading Area
See Technology section for more information on CE
STR genotyping is performed by comparison of sample data to allelic ladders
STR Allele Frequencies
TH01 Marker
*Proc. Int. Sym. Hum. ID (Promega) 1997, p. 34
FBI’s CODIS DNA Database
Combined DNA Index System
Used for linking serial crimes and unsolved cases with repeat offenders
Launched October 1998
Links all 50 states
Requires >4 RFLP markers
and/or 13 core STR markers
Current backlog of >600,000 samples

DNA is Only One Piece
of the Crime Solving Puzzle
DNA
SAFIS
Tox
Interviews
Technology Transition
“DNA Fingerprinting” – Dr. Alec Jeffreys – 1985
“DNA Profiling” – FBI (RFLP) – 1988
PCR DQalpha & Polymarker– 1990
PCR STRs – 1993
Mitochondrial DNA - 1996
Multi-plex STRs – 1997
Y-chromosome analysis
SNPs
Chips
Factors Driving the Changes
Discrimination potential
Sensitivity
Reliability
Ease of interpretation
Speed of analysis
Cost
Standards Development
Technical Working Group (TWGDAM)
Federal DNA Advisory Board (DAB)
Quality assurance standards
NDIS Standards
Data compatibility
FBI approved test methods
FBI approved test kits
Accessible only by authorized users for criminal justice identification purposes
Controls on data permitted in system
Strict controls on access and dissemination
Periodic audits
Data Sharing & Standards
CODIS (Combined DNA Index System)
Convicted Offender Index
Forensic Index
Missing & Unidentified Persons
Data compatibility (13 “core loci” STRs)
Data integrity (quality assurance standards)
Data transfer, retention and access controls
CODIS; the National DNA Database
Over 1 million DNA profiles of convicted offenders
Over 36,000 forensic profiles
Participation by 42 States
Over 5000 investigations aided
13 CODIS Core STR Loci with Chromosomal Positions
CSF1PO
D5S818
D21S11
TH01
TPOX
D13S317
D7S820
D16S539
D18S51
D8S1179
D3S1358
FGA
VWA
AMEL
AMEL