Works at indian river crime lab in ft. pierce, Fl
Cover “treasure coast”
A serologist
Walker’s background
Graduated UCF in 2007
BS in forensics (biochem) & minor in chem and micro bio
Interned at FDLE in drug chem section
Went forensic tech -> analyst
Cocaine is branded
What does a forensic biologist do?
Examine evidence to find bio material, then develop DNA profile from any bio evidence found
Sources of biological evidence
Blood, semen, saliva, urine, hair, teeth, bone, tissue, epithelial cells
Limitations of DNA testing
Temperature, sunlight, humidity
Age of stain
Contamination
Material stain is on
Conventional Serology
Identify body fluids through visual exam-> presumptive testing -> confirmatory testing
Small amounts of sample are typically used in analysis b/c you should only use as much as you need so you have some leftover
In military, if you're going to use all of the sample you have to tell prosecution & defense attorneys so they can come and witness you use all the evidence
Evaluate hair for DNA potential
Tests for blood
Presumptive: Kastle-meyer
Take a piece of filter paper and rub it on blood, if it turns pink positive- if not, negative
Confirmatory: Hematrace
Used to detect if blood is human (or ferret,,,?)
Only done as an extra step to check
Tests for Semen
Presumptive: acid phosphatase
Place filter paper and watch colour again, same as kastle meyer but turns red
Requires confirmatory test as many things can cause false positive (pre pubescent vaginal fluid, pregnant women vaginal fluid, feminine products)
Confirmatory: p-30
Microscopic exam
Tests for saliva
Presumptive: amylase plate
Agarose gel with starch tests for presence of alpha amylase from saliva
Also presumptive: SALIgAE
Proprietary…
Saliva makes it turn yellow
Tests for hair
Presumptive: visual examination
Look at medulla to determine if its human- ours is amorphous, animals have other distinct patterns
Anagen hair is suitable for DNA (has lots of pigment, attached at the root)
Telogen hair is NOT suitable for DNA (not much pigment, not attached at the root; already dead and removed itself from your root)
What is DNA?
Double helix that makes up your genome
Compressed into chromosomes
23 pairs found on humans, with the last pair being made up of a chromosome from each parent
Why do we use it?
Same in all cell types
Two copies: one from each parent
Stays the same through your whole life and is unique to everyone
Stable molecule
Types of DNA testing
Nuclear testing (STR & PCR based)
Mitochondrial testing (maternally inherited)
Y-STR testing (on y chromosome, paternally inherited)
STRs
2-7 BP long, side by side repeats
Same thing over and over
CAT CAT CAT CAT
21 assigned loci; don't tell you about person’s genetics
DNA Analysis
Stain is ID’d
Cutting or swabbing taken placed into tube
5 step process begins:
1. Extraction
Add heat and silica beads to attract DNA, press silica beads to magnet once DNA is attached
Can be done either:
Manually
Robotics (B3K/automated)
DNA IQ
Prepfiler
Differential extraction done to extract sperm specifically, making two samples from one
2. Quantification
Estimate of amount of DNA in sample
Useful in quality control and analysis
Done through Real time PCR (qPCR)
3. Amplification
Done with PCR!
Repeated 29 times
Done with commercial kits like Globalfiler & Powerplex fusiona
4. Separation
5. Analysis of Data
1 nanogram of DNA is ideal for amplication
Three possible outcomes:
Inclusion
Individual cannot be excluded as a potential contributor of DNA to evidence sample
Exclusion
Individual is eliminated as a potential contributor of DNA to the evidence sample
Inconclusive
DNA profiling results are not detailed enough to know or not
Once person is included…
Apply likelihood ratio: LR= Pr(E/Hp)/ Pr (E/ Hd)
E= evidence DNA profile
Hp= prosecutors hypothesis- DNA belongs to POI
Hd= defense hypothesis- DNA belongs to someone else
If LR> 1, Hp is supported
If LR<1, Hd is supported
If Lr=1, it doesn't support either
Ex.
E is 15,15
POI is 15,15
Frequency of 15 allele is .3
Remember hardy-weinberg (for homozygote, p^2+p(1-p) )
Applying formula: LR= 1/ hardy weinberg= 1/.0963= 10.38
Bc the evidence profile is approximately 10 times more likely to be observed if it originated from POI than if it originated from someone else.
Probabilistic Genotyping
Use bio modeling, statistical theory, computer algorithms, & probability distributions to infer genotypes & calculate LRs
Genotype weights are primary output
Multiple PG software available
Ex. STRmix
Paternity Statistics
For one obligate paternal allele: 2p-p^2
For two: 2(p+q)-(p+q)^2
Parentage index is approximately 1 billion
CODIS
FBI established DNA database containing DNA profiles from convicted offenders, forensic crime scene samples & missing persons
Allows searches of profiles between states
Want to look at some cases?
Case #1
Store owner robbed and tied up, suspects were wearing blue masks that they dropped on the way out
Masks provided techs with full DNA profiles of each suspect
Case #2
Stolen vehicle- steering wheel swabbed
Provided 4 person mix
Case #3
Multiple burglaries with similar MO
First time burglarizing occupied home
Pizza box found on counter
Suspect took a bite from a piece and then left the crust
Suspect ID’d who was near scene of multiple of the burglaries
Case #1: Bullet Lead Analysis
First used after kennedy assasination
Used chemistry to link CSI bullets to those posessed by suspects
Theory: each batch of lead had a unique elemental makeup
In 2004, ruled unreliable and misleading b/c of how overstated connection between bullets was
FBI responded by saying “sir yes sir”, but never went back to remedy any cases that may have been a mistrial bc of inclusion of bullet lead analysis as evidence
Brought to light by metallurgist William Tobin, a whistleblower
Found that bullets made in the same batch did not always match
In 2007, identified 250 cases of concern (12 of which were reversed)
FBI finally admitted to proposing erroneous testimony
FBI used faulty statistical calculations to declare bullets matched even when they were slightly off, and overstated significance of match
Basically claimed that if error bars were close enough to overlapping, should be considered statistically similar
Case #2: Microscopic Hair Comparison
Another flawed FBI testimony!
Found 26/28 of examiners overstated comparison match done by hair examiners
Came to light from a Washington Post Report
FBI did acknowledge immediately that there was an error, and that they didn’t consider confirmation bias, prosecutorial bias and outcome bias
But claimed it wasnt their fault because there wasnt enough hair research